比特派最新版本下载|bitter flavor _比特派(Bitpie)官网-比特派钱包app官方下载-bitpie官网下载app

BITTER中文(简体)翻译：剑桥词典

词典

翻译

语法

同义词词典

+Plus

剑桥词典+Plus

Shop

剑桥词典+Plus

我的主页

+Plus 帮助

退出

剑桥词典+Plus

我的主页

+Plus 帮助

退出

中文 (简体)

查找

英语-中文（简体）

bitter 在英语-中文（简体）词典中的翻译

bitteradjective uk

Your browser doesn't support HTML5 audio

/ˈbɪt.ər/ us

Your browser doesn't support HTML5 audio

/ˈbɪt̬.ɚ/

bitter adjective

(ANGRY)

Add to word list

B2 Someone who is bitter is angry and unhappy because they cannot forget bad things that happened in the past.

极为不满的；怨恨的；无法释怀的

I feel very bitter about my childhood and all that I went through.

我对童年和曾经历的一切感到怨恨。

She'd suffered terribly over the years but it hadn't made her bitter.

多年来她受尽了苦，可她并不怨恨。

B2 A bitter experience causes deep pain or anger.

使人痛苦的；令人愤怒的

Failing the final exams was a bitter disappointment for me.

期末考试不及格令我非常痛苦。

She learned through bitter experience that he was not to be trusted.

她从痛苦的经历中意识到他不可信。

B2 expressing a lot of hate and anger

充满仇恨的；愤怒的；激烈的

a bitter fight/argument

激烈的战斗／争吵

bitter recriminations

激烈的反诉

He gave me a bitter look.

他狠狠地瞪了我一眼。

更多范例减少例句They had a bitter quarrel over some money three years ago and they haven't spoken to each other since.They were bitter foes for many years.a bitter disputeShe cried bitter tears when she got the letter.They had a bitter, messy divorce.

bitter adjective

(TASTE)

B1 with an unpleasantly sharp taste

苦的，有苦味的

a bitter flavour/taste/liquid

苦味／苦味液体

更多范例减少例句Radicchio has a slightly bitter flavour.The tip of the tongue is sensitive to salt and sweet stimuli and the back of the tongue is sensitive to bitter stimuli.In today's lesson we'll look at the four types of tastes - sweet, salty, bitter and sour.The additive quinine, which you find in tonic water, is very bitter.They add syrup to the medicine to try and mask the bitter taste.

bitter adjective

(COLD)

B2 Bitter weather is extremely cold, especially in a way that causes physical pain.

（天气）严寒的，刺骨的

a bitter wind

刺骨的寒风

Wrap up warmly - it's bitter outside.

穿暖和些——外面特别冷。

同义词

Arctic figurative

biting (COLD)

freezing

glacial (ICE/COLD)

习语

the bitter fruits of something

a bitter pill (to swallow)

to the bitter end

bitternoun [ U ] uk

Your browser doesn't support HTML5 audio

/ˈbɪt.ər/ us

Your browser doesn't support HTML5 audio

/ˈbɪt̬.ɚ/

UK a type of dark brown beer with a bitter taste

苦啤酒

a pint of bitter

一品脱苦啤酒

比较

mild noun

bitters

a strong, bitter alcoholic drink made from spices and plant products that is mixed with other alcoholic drinks: Why not use different bitters to spice up classic drinks?

(bitter在剑桥英语-中文（简体）词典的翻译 © Cambridge University Press)

bitter的例句

bitter

This was a dispute that was as bitter as it was pointless, because ultimately insoluble.

来自 Cambridge English Corpus

They were not bitter exaggerations by frustrated entrepreneurs, but plausible explanations of why ventures had collapsed.

来自 Cambridge English Corpus

In the face of a violent liberation movement, they would probably have fought to the bitter end.

来自 Cambridge English Corpus

This definition at karela may send one scurrying to balsam pear and bitter gourd for clarification or confirmation.

来自 Cambridge English Corpus

The reactions to the demolition were strong and bitter.

来自 Cambridge English Corpus

Instead, his narrative ends with a bitter truism: everyone is alone, whether in captivity or liberty.

来自 Cambridge English Corpus

The boy is angry and bitter and wants his fake father to stay in jail.

来自 Cambridge English Corpus

When convicted, there was only so much they could do to prevent a sentence of exemplary harshness and a bitter gaol experience.

来自 Cambridge English Corpus

示例中的观点不代表剑桥词典编辑、剑桥大学出版社和其许可证颁发者的观点。

B2,B2,B2,B1,B2

bitter的翻译

中文（繁体）

愩怒，怒氣, 極為不滿的, 怨恨的…

查看更多内容

西班牙语

resentido, amargo, de resentimiento…

查看更多内容

葡萄牙语

amargo, ressentido, penoso…

查看更多内容

更多语言

in Marathi

日语

土耳其语

法语

加泰罗尼亚语

in Dutch

in Tamil

in Hindi

in Gujarati

丹麦语

in Swedish

马来语

德语

挪威语

in Urdu

in Ukrainian

俄语

in Telugu

阿拉伯语

in Bengali

捷克语

印尼语

泰语

越南语

波兰语

韩语

意大利语

कटू, भूतकाळातील काही वाईट घटना विसरता न आल्याने मनात राग आणि दुःख असण्याची भावना, कडवट…

查看更多内容

苦い, 苦味のある, 腹を立てている…

查看更多内容

kızgın, içerlemiş, çok kötü…

查看更多内容

amer/amère, âpre, violent/-ente…

查看更多内容

amarg, ressentit, glacial…

查看更多内容

bitter, zuur, snijdend…

查看更多内容

ஒருவர் கோபமாகவும், மகிழ்ச்சியற்றவராகவும் இருக்கிறார், ஏனென்றால் கடந்த காலத்தில் நடந்த மோசமான மற்றும் கசப்பான விஷயங்களை அவர்களால் மறக்க முடியாததால்.…

查看更多内容

(स्वभाव में) कड़वाहट, द्वेषपूर्ण, (स्वाद का) कड़वा…

查看更多内容

(પ્રકૃતિ માં) કડવું, દ્વેષપૂર્ણ, ઘણો નફરત અને ગુસ્સો વ્યક્ત કરવો…

查看更多内容

bitter, bidende…

查看更多内容

bitter, besk, förbittrad…

查看更多内容

pahit, pengalaman pahit, penuh kebencian…

查看更多内容

bitter, bitterkalt…

查看更多内容

bitter, besk, forbitret…

查看更多内容

کڑواہٹ, تلخی, شدیدغصہ اور ناراضگی…

查看更多内容

гіркий, гострий, запеклий…

查看更多内容

разочарованный, обиженный, злой…

查看更多内容

చేదు, కటువైన, పుల్లని…

查看更多内容

مُرّ, يَشْعُر بالمَرارة, قارِس…

查看更多内容

তিক্ত, কোনো ব্যক্তি তিক্ত, রাগান্বিত এবং অসন্তুষ্ট কারণ অতীতে ঘটে যাওয়া খারাপ ঘটনাগুলি ভুলতে পারে না।…

查看更多内容

kyselý, hořký, drsný…

查看更多内容

pahit, getir, penuh kebencian…

查看更多内容

ขม, ขมขื่น, ซึ่งเป็นศัตรู…

查看更多内容

đắng, cay đắng, gay gắt…

查看更多内容

rozgoryczony, zaciekły, zawzięty…

查看更多内容

쓴, 화가난, 혹독히 추운…

查看更多内容

amaro, risentito, (freddo) pungente…

查看更多内容

需要一个翻译器吗？

获得快速、免费的翻译！

翻译器工具

bitter的发音是什么？

在英语词典中查看 bitter 的释义

浏览

bits and bobs phrase

bits and pieces idiom

bitstream

bitten

bitter

bitter chocolate

bitter lemon

bitterly

bitterly cold phrase

bitter更多的中文(简体)翻译

全部

bitter lemon

bitter chocolate

a bitter pill (to swallow) idiom

to the bitter end idiom

the bitter fruits of something idiom

查看全部意思»

惯用语

a bitter pill (to swallow) idiom

to the bitter end idiom

the bitter fruits of something idiom

查看全部惯用语意思»

“每日一词”

white chocolate

a sweet, cream-coloured food made from cocoa butter, sugar, and milk, that is usually sold in a block

关于这个

博客

Renowned and celebrated (Words meaning ‘famous’)

March 13, 2024

新词

inverse vaccine

March 11, 2024

已添加至 list

回到页面顶端

内容

英语-中文（简体）例句翻译

©剑桥大学出版社与评估2024

学习

新词

帮助

纸质书出版

Word of the Year 2021

Word of the Year 2022

Word of the Year 2023

开发

词典API

双击查看

搜索Widgets

执照数据

关于

无障碍阅读

剑桥英语教学

剑桥大学出版社与评估

授权管理

Cookies与隐私保护

语料库

使用条款

京ICP备14002226号-2

©剑桥大学出版社与评估2024

剑桥词典+Plus

我的主页

+Plus 帮助

退出

词典

定义

清晰解释自然的书面和口头英语

英语

学习词典

基础英式英语

基础美式英语

翻译

点击箭头改变翻译方向。

双语词典

英语-中文（简体）

Chinese (Simplified)–English

英语-中文（繁体）

Chinese (Traditional)–English

英语-荷兰语

荷兰语-英语

英语-法语

法语-英语

英语-德语

德语-英语

英语-印尼语

印尼语-英语

英语-意大利语

意大利语-英语

英语-日语

日语-英语

英语-挪威语

挪威语-英语

英语-波兰语

波兰语-英语

英语-葡萄牙语

葡萄牙语-英语

英语-西班牙语

西班牙语-英语

English–Swedish

Swedish–English

半双语词典

英语-阿拉伯语

英语-孟加拉语

英语-加泰罗尼亚语

英语-捷克语

英语-丹麦语

English–Gujarati

英语-印地语

英语-韩语

英语-马来语

英语-马拉地语

英语-俄语

English–Tamil

English–Telugu

英语-泰语

英语-土耳其语

英语-乌克兰语

English–Urdu

英语-越南语

翻译

语法

同义词词典

Pronunciation

剑桥词典+Plus

Shop

剑桥词典+Plus

我的主页

+Plus 帮助

退出

中文 (简体)

Change

English (UK)

English (US)

Español

Русский

Português

Deutsch

Français

Italiano

中文 (简体)

正體中文 (繁體)

Polski

한국어

Türkçe

日本語

Tiếng Việt

हिंदी

தமிழ்

తెలుగు

关注我们

选择一本词典

What Is Bitterness?

button

SUBSCRIBE TO OUR NEWSLETTERS

Email Address

There was an error. Please try again.

Close search

Recipes

Breakfast & Brunch

Lunch

Appetizers & Snacks

Dinner

Desserts

Side Dishes

Breads

Cocktails

View all

By Region

American Food

Asian Food

European Food

Latin American Food

Middle Eastern Food

African Food

Australian Food

View all

Ingredients

Chicken Recipes

Beef Recipes

Pork Recipes

Fish & Seafood

Fruit & Veggie Recipes

Cheese Recipes

View all

Occasions

St. Patrick's Day

Passover

Easter

Soup Season

Winter Recipes

Seasonal Produce

View all

How-Tos

Learning How to Cook

Cooking Techniques & Tips

Cooking Equipment

Knife Skills

Ingredients

What to Buy

View all

What to Buy

How We Test Products

Small Appliances

Cookware

Grills & Grilling

Coffee & Tea

Kitchen Gadgets

Storage & Organization

Bakeware

This Is Fire!

View all

News

The Rice Package

Voices

Trends

View all

About us

About Us

Editorial Guidelines

Anti-Racism Pledge

Newsletters

Cooking School

Learning How to Cook

Cooking FAQs

What Is Bitterness?

A Guide to Understanding Bitterness in Foods

Bethany Moncel

Professional blogger and cookbook author Bethany Moncel has become an expert on making delicious, healthy meals on a budget. She also holds a nutritional science degree.

Learn about The Spruce Eats'

Editorial Process

Updated on 09/20/22

Site Currently Unavailable

400_401_403_bgr

dots

site-currently-unavailable

Site Currently Unavailable

If you are the owner, contact your hosting provider for more information.

clouds_shape

Page restricted | ScienceDirect

Your Browser is out of date.

Update your browser to view ScienceDirect.

View recommended browsers.

Request details:

Request ID: 863cdebd1ee71248-HKG

IP: 49.157.13.121

UTC time: 2024-03-13T14:56:20+00:00

Browser: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/97.0.4692.71 Safari/537.36

About ScienceDirect

Shopping cart

Contact and support

Terms and conditions

Cookies are used by this site. By continuing you agree to the use of cookies.

Copyright Â© 2024 Elsevier B.V., its licensors, and contributors. All rights are reserved, including those for text and data mining, AI training, and similar technologies. For all open access content, the Creative Commons licensing terms apply.

Bitter or not? BitterPredict, a tool for predicting taste from chemical structure | Scientific Reports

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain

the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in

Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles

and JavaScript.

View all journals

Explore content

About the journal

Publish with us

RSS feed

nature

scientific reports

articles

article

Bitter or not? BitterPredict, a tool for predicting taste from chemical structure

Download PDF

Article

Open access

Published: 21 September 2017

Bitter or not? BitterPredict, a tool for predicting taste from chemical structure

Ayana Dagan-Wiener1,2, Ido Nissim1,2, Natalie Ben Abu1,2, Gigliola Borgonovo3, Angela Bassoli

ORCID: orcid.org/0000-0001-5442-38083 & …Masha Y. Niv1,2 Show authors

Scientific Reports

volume 7, Article number: 12074 (2017)

Cite this article

17k Accesses

99 Citations

27 Altmetric

Metrics details

Subjects

Machine learningVirtual drug screening

AbstractBitter taste is an innately aversive taste modality that is considered to protect animals from consuming toxic compounds. Yet, bitterness is not always noxious and some bitter compounds have beneficial effects on health. Hundreds of bitter compounds were reported (and are accessible via the BitterDB http://bitterdb.agri.huji.ac.il/dbbitter.php), but numerous additional bitter molecules are still unknown. The dramatic chemical diversity of bitterants makes bitterness prediction a difficult task. Here we present a machine learning classifier, BitterPredict, which predicts whether a compound is bitter or not, based on its chemical structure. BitterDB was used as the positive set, and non-bitter molecules were gathered from literature to create the negative set. Adaptive Boosting (AdaBoost), based on decision trees machine-learning algorithm was applied to molecules that were represented using physicochemical and ADME/Tox descriptors. BitterPredict correctly classifies over 80% of the compounds in the hold-out test set, and 70–90% of the compounds in three independent external sets and in sensory test validation, providing a quick and reliable tool for classifying large sets of compounds into bitter and non-bitter groups. BitterPredict suggests that about 40% of random molecules, and a large portion (66%) of clinical and experimental drugs, and of natural products (77%) are bitter.

Similar content being viewed by others

Bumblebees socially learn behaviour too complex to innovate alone

Article

Open access

06 March 2024

Alice D. Bridges, Amanda Royka, … Lars Chittka

Microdosing with psilocybin mushrooms: a double-blind placebo-controlled study

Article

Open access

02 August 2022

Federico Cavanna, Stephanie Muller, … Enzo Tagliazucchi

Identifying general reaction conditions by bandit optimization

Article

28 February 2024

Jason Y. Wang, Jason M. Stevens, … Abigail G. Doyle

IntroductionBitter taste is a basic taste modality, which is believed to have evolved to protect animals from consuming toxic food1. It was estimated that the number of bitter compounds is in the thousands or tens of thousands2. We have established BitterDB http://bitterdb.agri.huji.ac.il/dbbitter.php, a database which now holds about 700 compounds that were reported to have bitter taste or to activate at least one human bitter receptor in cell-based assays3. BitterDB includes structurally diverse compounds such as ions, peptides, alkaloids, polyphenols, glucosinolates and more. The full repertoire of molecules that activate bitter taste receptors is currently unknown. Furthermore, the size of the bitter chemical space, the abundance of bitter compounds in chemical space (both natural and synthetic) are currently unknown.The bitter taste is recognized in human by 25 G-protein-coupled receptors (called T2Rs or TAS2Rs). These receptors are expressed in the oral cavity, as well as in the gastrointestinal tract, the upper airways, the heart and in additional tissues4,5,6. Indeed, T2Rs have physiological roles in the digestive process7, affecting respiration and activating the immune system8,9. T2Rs have been suggested as novel therapeutic targets for asthma10 and respiratory infections8,11.Prediction of bitterness has therefore several practical implications: predicting bitter compounds within the human metabolome may suggest yet unknown endogenous ligands of T2Rs. Identifying bitterness of approved drugs may be useful for their repurposing for novel indications12. On the other hand, bitterness of drug molecules presents compliance problems13,14 and early flagging of potential bitterness of a drug candidate may help its further development. Bitterness prediction is also important for the food industry, for example indicating if key ingredients are likely to be bitter and therefore require application of masking procedures15,16.The main challenge in computational bitterness prediction arises from the chemical diversity of bitter compounds, while only minor differences between bitter and non-bitter compounds exist in some cases17. Therefore, most of the predictive models focused on specific families of bitter compounds18 such as cynaropicrin and grosheimin derivatives19,20, and benzenesulfamates21. Bitterness of peptides was also examined in several past studies22,23.Several successful ligand-based and structure-based approaches predict bitter molecules that activate a specific bitter receptor and were successful in predicting ligands of bitter receptors with a large number of known ligands24,25,26.Methods for predicting bitterness using machine learning approaches were developed to address bitterness prediction not limited to a particular T2R or chemical family: Rodgers and colleagues used a dataset of 649 bitter and 13,530 randomly selected molecules which were approximated as non-bitters, to develop a Naïve Bayes classifier based on circular fingerprints (MOLPRINT2D27) and information-gain feature selection16. The proprietary classifier and dataset are unavailable to the research community.Huang et al.28 used support vector machine (SVM) algorithm to build BitterX tool, which is based on physicochemical descriptors of the compounds and can be used to predict bitterness. In combination with receptors descriptors it predicts also which of the T2Rs the compound may activate. The positive set was built from BitterDB3 and additional manually curated compounds (~500 in total); the negative set was comprised mainly from representative structures of compounds which were not described as ‘bitter’ in the Available Chemicals Directory (ACD, http://www.accelrys.com). BitterX is available via http://mdl.shsmu.edu.cn/BitterX/.In the current study we developed a general predictor applicable to diverse chemical families, to allow answering the following questions: I) For any given compound, is the compound likely to be bitter or non-bitter? II) In a library of compounds (i.e. food-derived, drug-related, natural, etc.) – what percent of compounds is expected to be bitter? In other words, what is the abundance of bitter compounds in a given chemical space? Finally – III) Are there characteristic properties of bitter vs. non-bitter compounds?Below we describe the development and applications of the machine learning method BitterPredict that addresses these questions.ResultsA key step in development of machine learning predictor is the construction of true positive and true negative datasets. Below we describe these datasets and analyze their properties in comparison to the ChEBI29 database, which is taken as a set of random molecules.Chemical Entities of Biological Interest (ChEBI) setA dataset of tens of thousands ‘small molecular entities’. ChEBI stands for ‘Chemical Entities of Biological Interest’. It is a freely available database of ‘small molecular entities’, developed at the European Bioinformatics Institute (EBI)29. The compounds and other molecular entities within this dataset are either products of nature or synthetic products used to intervene in the processes of living organisms. After data curation, 41,132 molecules were ready for prediction with 29,340 molecules in Bitter Domain. This set was used as “random” set, aiming representing a general chemical space.The positive (bitter) setThe major resource for the positive set includes 632 molecules from BitterDB3 and additional 59 compounds from a recent study by Rojas at el.30.691 (97%) of the positive set molecules have molecular weight (MW) < = 700 and hydrophobicity (AlogP) range −3 < = AlogP < = 7 (Fig. 1). Therefore, the predictive model was restricted to this “bounding box”31. Molecules outside this domain can be assumed to be non-bitter, unless a more specialized approach (such as ligand-based approach trained on the specific chemical family) suggests otherwise. We named this applicability domain “Bitter Domain”. All datasets used for training and prediction were first filtered to exclude molecules outside this domain.Figure 1“Bitter Domain”: scatter plot of AlogP vs MW of the random molecules and the positive set molecules. The green rectangle represents the Bitter Domain which is defined by −3 =

The negative setTo the best of our knowledge, no publically available set of non-bitter molecules exists, and such data is not easy to retrieve. Previous studies used random molecules16 or undisclosed in-house sets28. We established a dataset of ~2,000 non-bitter molecules (all within the Bitter Domain), composed from three subsets: non-bitter flavors, sweet molecules, and tasteless molecules, as detailed below.Non-bitter flavors subset: The non-bitter flavors subset comprises 1,360 ‘likely not bitter’ molecules that were collected from Fenaroli’s Handbook of Flavor Ingredients32 in an automated fashion. Compounds were considered as non-bitter if the word bitter did not appear in its description. In addition to flavors, this subset also contains 56 molecules which were cited in TOXNET33 as tasteless and 35 additional non-bitter molecules that were manually extracted from the literature17,34,35.Sweet and tasteless subsets were built from sweet (not reported to be bitter) (336) and tasteless (130) compounds recently reported by Rojas at el.30.Random moleculesRandom molecules were represented by molecules from Chemical Entities of Biological Interest (ChEBI) dataset29 (see Methods). The set includes more than 40,000 molecules, among them almost 30,000 compounds within the Bitter Domain. This set was used as reference set, representing a general and wide chemical space.Principal component analysis (PCA)PCA using 12 basic physicochemical properties of the negative set, positive set and random molecules within the Bitter Domain, shows that the negative set covers almost all of Bitter Domain, and that each negative subset (flavors, sweet, and tasteless) captures a partial region of the chemical space in the Bitter Domain (Fig. 2).Figure 2Training data chemical space (a) PCA of the negative set, positive set and random molecules within the Bitter Domain (for PCA details see Methods). The bitter molecules (green) spread widely inside the Bitter Domain. Each non-bitter sub set covers distinct sub-space; however the combined non-bitter set covers almost all the domain, though not uniformly distributed. Principle component 1 (PC1) and PC2 explain ~60% and ~17% of the variation, respectively. (b) Examples of molecule structures from the positive set and the different negative sub sets.Full size image

Preliminary testing indicated that the negative set has a dominant effect on the quality of prediction (Supplementary Table S2 shows prediction results using only the non-bitter flavors as the negative set, causing over-prediction of bitterness). The low variance in properties of the non-bitter flavor subset (see Supplementary Fig. S1 for histograms) is likely due to the fact that the majority of molecules in this set are volatile molecules, most of which are relatively small.Molecular descriptorsDue to the widely assumed connection between bitterness and toxicity36, ADME/Tox (absorption, distribution, metabolism, excretion and toxicity) descriptors from the QikProp package (version 4.6 Schrödinger, LLC, New York, NY, 2015) were used. The QikProp package predicts physically and pharmaceutically significant properties of organic molecules based on the full 3D molecular structure. Ten additional basic physiochemical descriptors which are not part of the QikProp package were also used (see Methods for details).Bivariate statistical analysis of the training data with selected QikProp descriptors indicated combined properties ranges which are enriched with bitter molecules, such as medium-low skin permeability (QPlogKp) and medium-high hydrophobic component of total solvent accessible surface area (FOSA). The analysis also indicated that bitter molecules are predicted to have low MDCK cell permeability and low predicted brain/blood partition coefficient (QPlogBB) compared to the non-bitter molecules (some of the data is shown in in Supplementary Fig. S2), supporting the idea that QikProp properties are useful for the classifier.AlgorithmsAdaBoost with decision trees as weak learners37 is an adaptive ensemble method, in which the decision trees are built sequentially, by learning from mis-classified samples of the former decision tree. The different sizes of the bitter and non-bitter datasets, as well as different sizes of subsets within the non-bitter, were balanced via the initial observation weight vector (see Methods for details).Model performanceIn order to avoid overfitting and to get a clear picture of the model performance, the input data was divided randomly to 70% training set and to 30% hold out test set. This split ensures the original proportions in both the training and the test sets. Several parameters were adjusted in respect to the test set performance (for details see Methods).The classification performance parameters on training and test sets are listed in Table 1. The sensitivity and specificity on the training set were 0.91 and 0.94 respectively and on the test set 0.77 and 0.86 respectively. These results correspond to score threshold of zero (samples with prediction scores larger than zero are classified as bitter and samples with score smaller than zero as non-bitter). The performance on the subsets of the non-bitter molecules is also listed in Table 1, suggesting average specificity >0.8. This analysis shows that the model offers a good classification to bitter and non-bitter molecules.Table 1 Results on training and test sets.Full size table

In some cases the goal may be obtaining high sensitivity, while specificity is less important; in other cases higher specificity is most needed. Changing the threshold score which determines bitter/non-bitter decision can be used to fine-tune the predictor for the needed purpose. Test set compounds with prediction scores greater than 0.6 lead to false positive rate (FP/(FP + TN)) lower than 0.05 and sensitivity above 0.5. Prediction scores less than −0.7 lead to false negative rate (FN/(FN + TP)) lower than 0.01 and specificity above 0.5 (see Fig. 3). Therefore, we suggest a cutoff score > 0.6 when high confidence bitter predictions are required, and a cutoff score of <−0.7 for high confidence non-bitter predictions.Figure 3AdaBoost prediction score distribution on the test set: histogram showing the prediction score distribution on the test set. Dashed lines indicate thresholds for more reliable predictions; above score 0.6 and beneath −0.7 false positive and false negative rates respectively, are low.Full size image

ValidationNext, BitterPredict performance was evaluated via three approaches:I) Validation using external sets; II) Validation by literature mining; and III) Validation by taste tests.I) External sets. In principle, test set overfitting can be caused by selecting the models with the best prediction statistics on the hold out test set38. In order to better assess the classifier performance, several external (new) sets were gathered.Bitter New(molecules collected from the literature and not included in BitterDB or the training set): 29 molecules not included in the training or test sets, were collected from six publications39,40,41,42,43,44. 6 of the molecules lay outside Bitter Domain (MW > 700), 5 of those are tannin compounds (procyanidin B, C and PGG).The prediction was performed on the remaining 23 structurally diverse compounds (MW ranges from 60 to 600 (g/mol) and AlogP ranges from −1.5 to 6).BitterPredict correctly predicted 17 out of the 23 molecules (0.74 Sensitivity). 14 of the 17 true positive molecules got prediction scores higher than 0.6. 3 of the 6 bitter molecules which were misclassified as non-bitter are ethyl esters, and only one of the six misclassified molecules has a high negative score (<−0.7).UNIMI setPreviously unpublished data from Bassoli lab at the University of Milan (UNIMI). The set contains 64 molecules, including 23 bitter, 33 non-bitter, 4 “unpleasant” and 4 with undefined taste. The “unpleasant” molecules were excluded due to the difficulty to discriminate between bitter and other unpleasant tastes.This set is very challenging, because it contains molecules which share the same scaffold but elicit different tastes. BitterPredict was correct on 78% of the bitter and 85% of the non-bitter compounds in this set. In the misclassified samples there are 2 molecules for which stereoisomers are present in this set, but with different taste (see Supplementary Fig. S3), exemplifying a case which might be better addressed by specialized predictors.Molecules from the Phytochemical Dictionary54 bitter and 39 non bitter (mostly sweet) molecules were extracted from the Phytochemical Dictionary book45 which includes information about taste of bioactive compounds from plants. These molecules were not part of the training or test sets. 6 of the 54 bitter molecules and 13 of the 39 non-bitter molecules were outside the Bitter Domain. BitterPredict correctly classified 98% of the bitter molecules and 69% of the non-bitter molecules. 14 of the 18 molecules correctly classified as non-bitter, scored <−0.7; 44 of the 48 molecules correctly classified as bitter, scored >0.6.Table 2 shows excellent results for the 3 unrelated external datasets, with sensitivity of 74%-98% and specificity of 69%-85%.Table 2 Prediction on external validation sets.Full size table

II) Validation by literature mining. We assumed that information about bitterness or other off-tastes of orally administered clinical drugs would be readily available in the literature or on the web.BitterPredict was applied to DrugBank set of FDA approved drugs46 and the compounds were sorted by their predicted bitterness score. The names of the top 30 compounds (most likely to be bitter) and bottom 30 compounds (most likely to be non-bitter) were submitted to datamining in scientific publications (using Google Scholar), chemical databases (PubChem47 and ChemSpider48) and in the web (using Google) with the word “taste” or “bitter taste” to get an indication of their taste.For the top predicted bitter compounds, 14 were described as bitter, 4 were indirectly described as bitter (for example the tablets that mainly include the compound of interest are described as bitter), 4 had description of unpleasant taste, for 8 no relevant data on taste could be found. Notably, none of the “predicted to be bitter” compounds had taste description other than bitter or unpleasant.For the top predicted non-bitter compounds: for 6 molecules (20%) notion that they are bitter or might be bitter was found. For 6 molecules (20%) taste description other than bitter (“tasteless”, “mint”, “sweet with bitter aftertaste” and “a known bitter masking agent”) was found. For the majority of the predicted to be non-bitter molecules (18 molecules, ~60%) no indication of taste could be found. Since bad taste of a drug is a typical complaint, we assume that lack of mention of bitter taste indicates these are not very likely to taste bitter.Overall, these results indicate that ~60% of top predicted to be bitter drugs were mentioned to have bitter taste, while only 20% of predicted to be non-bitter had potential mention of bitter taste (Fig. 4a).Figure 4BitterPredict Validation: (a) DrugBank literature-derived information: histogram showing the taste description found in datamining for the 30 most likely bitter compounds and the 30 most likely non bitter compounds according to BitterPredict predictions. (b) Sensory evaluation of compounds predicted to be non-bitter: Bars indicate mean ± s.e.m. The red horizontal line represents the mean of water bitterness (control). Asterisks indicate a significance difference (P < 0.05) from control by the two-tailed Dunnett test.Full size image

III) Validation by taste tests: To allow for sensory testing using human panel, BitterPredict was applied to Sigma-Aldrich Ingredients Catalog Flavors& Fragrances food which include Food-grade, Natural, Kosher or Halal materials. Among the 279 compounds extracted from Sigma-Aldrich flavors and fragrances catalog (http://go.sigmaaldrich.com/ff-catalog-download-safcglobal), only 14 were predicted bitter with score above 0.6. For 4 of these, molecules with the same name were already included in the BitterDB, and almost all the others were either insoluble, allowed for consumption only at very low concentration, or not readily available for purchase (see Supplementary file validation.xls for details), and were not tested. 105 compounds were predicted as non-bitter with score < −0.7. 6 compounds were chosen for sensory testing according to their prediction score, safety considerations and availability. The 6 non-bitter compounds were diluted in distilled water (0.5 mM) and tested in a sip&spit experiment by a panel of 12 subjects (see Methods for details). During the experiment the participants used nose clips to prevent smelling of solution odor. 5 out of the 6 tested compounds (83.3%) were rated equally or less bitter than the water used for dilution. Using the Dunnett test (alpha = 0.05) none of the 6 solutions differed significantly in bitterness from water (the significance for each of these comparisons was p > 0.8). Quinine, that was added to the experiment as an example of a known bitter compound, was significantly more bitter than water (the actual significance for this comparison was p < 0.0001) (Fig. 4b).In summary, all the validation approaches suggest a reliable and consistent performance of BitterPredict for both bitterness and non-bitterness prediction.Model interpretationIt is interesting to explore the contribution of the different descriptors to the model performance. This can be calculated (see Methods for details) from the contribution of the descriptor to reducing the error. Figure 5 shows the top 27% descriptors according to this analysis. The most important descriptor is the total charge. Indeed, the majority of the positively charged molecules are bitter molecules containing ammonium ion at physiological pH (Fig. 5b), suggesting that molecules with positively charged ammonium ion are more likely to be bitter than neutral or negatively charged ones.Figure 5Descriptors contribution to The AdaBoost model (a) 16 descriptors with the most significant contribution to the AdaBoost model (contribution score greater than 1*10−4). (b) Total charge distribution in the bitter and non-bitter sets.Full size image

Additional descriptors of importance are FOSA, FISA and PISA, which describe different parts of the molecular surface and are related to hydrophobicity.QplogHERG, QPlogKp, QPPMDCK, QplogKhsa, QPlogBB (blockage of Human ether-a-go-go-related gene (hERG) K+ channels, skin permeability, transport in the gut blood barrier, binding to human serum albumin and blood-brain barrier transport measure) are connected to potential toxicity of the molecule in the body, and were found to have a relatively large contribution to the prediction model. Bitter molecules are enriched (22%) with compounds which are predicted to block hERG K+ channels compared to the negative set (10%) (see Supplementary Fig. S5). This result is in accord with our recent observation24 that >20% of the bitter tastants in BitterDB are known inhibitors of the hERG K+ channels. It was recently found that bitter receptors are also expressed in heart tissue49, thus some compounds could be acting on both bitter taste receptors and hERG channel in the heart tissue. Interestingly, QikProp descriptors related to potential toxicity tend to have larger contributions to the prediction model than descriptors related to specific molecular property (such as number of ring atoms, number of carboxylic acid group etc.).Prospective predictionsThe BitterPredict classifier was next applied to several datasets of interest. Specifically, we used FooDB (http://foodb.ca/), a dataset of food constituents, DrugBank approved set46, natural product subset in ZINC15 database50 which holds commercially available secondary metabolites, and ChEBI29 as a representative set of random molecules. The predictions provide an estimate on the percentage of bitter molecules within related chemical datasets and are available via Supplementary files and via Ambinter website (http://www.ambinter.com/moleditor/web/), details in Data availability section below.The estimation of the percentage of predicted bitter molecules in the different datasets are detailed in Table 3.Table 3 Prospective predictions.Full size table

The prediction on ChEBI represents the estimation of abundance of bitter compounds among random molecules. Interestingly, BitterPredict suggests that ~43% of random molecules may have some bitter taste. This means that automatic assumption that a random molecule is tasteless may be wrong in many cases.Since ChEBI includes many synthetic molecules, it is of particular interest to evaluate the percentage of bitter compounds also among natural products. The natural products set used in this study contains secondary metabolites mainly from plants. Plants produce secondary metabolites including toxic chemicals as part of their defense system against herbivore attack, and many of these compounds are known to be bitter51,52. Indeed, 77% of the natural products library was predicted to have bitter taste. The high percentage of predicted bitter compounds in this set suggests that bitter taste may be among the most abundant tastes encountered in nature.Many drugs are well known to have bitter taste12,14,24. In accord with this notion, 66% of the DrugBank library were predicted to be bitter.Food ingredients represented in the FooDB, are predicted to include 38% bitter compounds. The relatively high percentage of bitter ingredients in FooDB is somewhat surprising, but may reflect the presence of many glucosinolates, terpenes, flavonoids and some alkaloids (such as quinine and caffeine) that are commonly consumed in foods and beverages such as coffee, tea, tonic water, vegetables and fruits, despite their bitterness.To conclude, BitterPredict suggests that for an arbitrary small molecule there is 40% chance that it would elicit at least some bitter taste. For a molecule that belongs to a set that is more related to bitterness (such as drugs and natural products) the chances for some bitterness are much higher, around ~70%.AccessibilityAccess to results and to BitterPredict software is detailed in the Data Availability section.DiscussionIn this study, we have extended the dataset of bitter molecules and established a dataset of 2000 non-bitter molecules. Using these datasets, we have developed a bitter/non-bitter classifier. The classifier’s performance was evaluated on several external sets, showing high and robust sensitivity and specificity (around 0.8). Furthermore, datamining of top predictions in DrugBank set of clinical drugs and sensory experiment on food and flavor ingredients further confirmed the excellent performance of the predictor.Application of BitterPredict to random molecules suggests that ~40% of the small molecules chemical space (ChEBI data set) have some bitter taste. Considering predictions with confidence score above 0.6 still predicts 11,115 (~30%) as bitter. This result is opposed to earlier assumption that random molecules are probably not bitter16.To study questions related to taste perception and its shaping through evolution53,54, natural compounds (rather than molecules synthetized by chemists during the last centuries) should be studied. Therefore, it is particularly interesting to estimate the percentage of bitter compounds among natural products. Our results suggest that above 77% of natural products have some bitterness.BitterPredict highlights the total charge, the surface related properties, and the potential toxicity of the molecule properties as important descriptors differentiating bitter from non-bitter. In former studies, the connection between bitterness and hydrophobicity of peptides was discussed and disputed23,55,56. In accord with previous studies20,57, hydrophobicity was not found to be a predictive feature of bitterness, with similar AlogP distribution of bitter and non-bitter sets within the Bitter Domain (see Supplementary Fig. S1). However, low values of AlogP (<−3) are rare in the positive set, while 25% of the sweet molecules set have AlogP < −3. This means that very hydrophilic compounds are unlikely to be bitter. In addition, there are differences in the distributions of the important features indirectly related to hydrophobicity such as FOSA and hydrophilic component of total solvent accessible surface area (FISA). The combination of properties lends the classifier the ability to discriminate between the bitter and the non-bitter molecules.The major advantages of BitterPredict is its high accuracy (~80%) and the ability to predict both bitterness and non-bitterness of a molecule based on its structure. The predictions are given a score, enabling the user to filter predictions according to intended use: for example such that false positive bitter predictions are rare, or such that false negative bitter predictions are rare. The relatively high sensitivity and specificity of BitterPredict enables exploration of large chemical spaces. The performance can be easily improved as more experimental data becomes available. The datasets established here are available to the users and will serve as a benchmark for further developments of bitterness prediction and classification methods.There are also some limitations of the method that should be kept in mind: the method is applicable only to compounds within the Bitter Domain; scarcity of available data on levels of bitterness does not yet allow the discrimination between strongly vs. weakly bitter compounds.Future studies will aim to classify the bitter space in further sub-categories in order to differentiate between strongly bitter vs. weakly bitter compounds, and between bitterness of compounds from different sources and habitats. The current analysis allows design of experiments on molecules outside of the current applicability domain, which will eventually lead to extension of the Bitter Domain.To the best of our knowledge, this is the first study that attempts to estimate the proportion of bitter molecules in a general chemical space and in specific chemical datasets. The possible applications of the BitterPredict classifier include studies of basic questions related to evolution of taste in different species, bitter taste receptors de-orphanization, and practical applications in food and drug development.MethodsData Sets preparationThe largest fragments of the chemical structures from all sets were uploaded to Maestro (Schrödinger Release 2015–4: MS Jaguar, Schrödinger, LLC, New York, NY, 2015). Salt ions, peptides, inconsistencies, and structures with less than 3 atoms were removed. 3D structures and protonation states at biological pH 7.0 ± 0.5 were generated with Epik and LigPrep (Schrödinger Release 2015-4: LigPrep, Epik, LLC, New York, NY, 2015) retained the original chirality of the compound (if specified). For each molecule, the conformer with the lowest energy was extracted. In case a molecule has two protonation states in the defined pH, both molecules were included. Molecules that cannot be neutralized were removed from this study due to QikProp descriptors calculation limitation. Duplicates (i.e molecules with identical values for each one of the descriptors used) were also removed.The sets were filtered to include only structures within the primary applicability domain, named, “Bitter Domain”: MW >700 and −3

Data sets used to build and analyze training and test sets:Bitter Set (positive): Include mainly molecules from BitterDB3, a database of almost 700 bitter compounds which were described as bitter in the literature, or were reported experimentally as capable of activating at least one human bitter taste receptor: 632 molecules from this database used for building the classifier. In addition 59 unique molecules from Rojas et al.30 study were added to the positive set.Non-Bitter Flavors: The dataset consists of 92 molecules which were cited in TOXNET33 as tasteless. 68 additional molecules were extracted from publications manually or using Marvin 6.1 naming capabilities, 2013, ChemAxon (http://www.chemaxon.com), 1,413 ‘probably not bitter’ molecules which were collected from Fenaroli’s handbook of Flavor ingredients32 in an automated fashion: compound was considered as non-bitter if the word bitter does not appear in its description fields. For these compounds the CAS numbers or generic names were manually extracted and used to look up the structures in PubChem47. Additional manual validation for the Fenaroli’s handbook molecules was performed by looking for the taste description of randomly chosen 30 molecules in other resources. We discovered only one molecule with unclear taste, and removed it. In addition, some amino acids were removed from this set due to conflicting description of their taste in different resources. Set size after data curation in Bitter Domain was 1,451.Sweet and Tasteless sets: 414 sweet compounds, and 130 tasteless compounds that were reported as sweet or tasteless in taste experiments, where extracted from recent study by Rojas et al.30. After removing duplicates with other non-bitter subsets and dataset preparation procedures the set sizes were 336 for sweet and 130 tasteless.Evaluation setsHold out sets that were not used in training and testing routines.Bitter New: 29 molecules that were collected recently from different publication39,40,41,42,43,44 to expand BitterDB and were not part of the train or test set. 23 molecules from this set that were in Bitter Domain were used for validation.UNIMI set: Molecules that were synthesized during years 1990 to 2000 as part of studies on taste active compounds, usually sweeteners. The compounds were submitted to preliminary tasting trials with a panel of 5–8 untrained panelists. The informed consent was insured by the responsible principal investigator (A. Bassoli, University of Milano), and the taste sessions were carried out following the common procedures at the time61.Phytochemical Dictionary:45 The dictionary includes 3,000 bioactive compounds from plants, for some of them a taste description is provided. The CAS numbers or generic names were manually extracted and used to look up the structures in PubChem47 or ChemSpider48. After data curation, 55 bitter and 39 non-bitter compounds were available, with 49 bitter and 26 non-bitter compounds in the Bitter Domain.Data set used for sensory evaluationSigma Ingredients Flavors & Fragrances: 1047 molecules where extracted from Sigma-Aldrich flavors and fragrances catalog (http://go.sigmaaldrich.com/ff-catalog-download-safcglobal) using Marvin 16.2.1 naming capabilities, 2016, ChemAxon (http://www.chemaxon.com). A large portion of the original data set (>50%) was identical to the Non-Bitter Flavors. After data curation, and removing compounds which are identical to compounds in the positive or negative set, 279 compounds were available, with 264 compounds in the Bitter Domain.Data set used for prospective predictionDrugBank approved46: The data set includes 1,621 Food and Drug Administration (FDA) approved small molecule drugs. After data curation, 1,553 molecules were ready for prediction, with 1,375 molecules inside Bitter Domain.FooDB: (http://foodb.ca/) A data set which holds tens of thousands of food constituent molecules. 24,399 molecules extracted from the FooDB SQL version. After data curation, 20,661 molecules where ready for prediction, with 13,588 molecules in Bitter Domain.Natural Products Dataset: A data set which holds 38,469 commercially available natural products and excludes the ZINC15 primary metabolites subset (was downloaded from ZINC1550). The data includes several datasets, among them are herbal and plants natural products sets58,59. After data curation 28,217 molecules where ready for prediction, with 27,474 molecules in Bitter Domain.PCAPCA was used in order to estimate the variance and chemical space of the datasets that were used in this study. Prior to the PCA analysis, the descriptors were normalized such that each descriptor has zero mean and unit variance. The PCA was performed using Matlab (version R2015a; Mathworks, Inc., MA, USA), using the 2D structures of the molecules and 12 basic physiochemical descriptors calculated with Canvas (Schrödinger Release 2015-4: Canvas, Schrödinger, LLC, New York, NY, 2015): molecular weight (MW), lipophilicity (ALogP, the atomic LogP), rotatable bonds count (RB), polar surface area (PSA), electrotopological states (estate), molecular refractivity (MR), molecular polarizability (Polar), hydrogen bond acceptor (HBA), hydrogen bond donner (HBD), rings count (ring), chiral centers count (chiral) and heavy atoms count (HA).DescriptorsIn addition to the 12 physicochemical descriptors detailed above, the total charge, and the number of aromatic rings descriptors were calculated with Canvas (Schrödinger Release 2015-4: Canvas, Schrödinger, LLC, New York, NY, 2015). 47 absorption, distribution, metabolism, and excretion (ADME/Tox) descriptors were calculated with QikProp module (Schrödinger Release 2015-4: QikProp, Schrödinger, LLC, New York, NY, 2015). The ‘predicted maximum transdermal transport rate descriptor’ (JM) from QikProp package was excluded due to high variance in values for very similar compounds, the HBA and HBD descriptors from the physicochemical descriptors were removed due to redundancy with QikProp descriptors. In total, 59 descriptors were used to build the model; the complete list of descriptors is available in Supplementary Table S3.Predictive modelsPreliminary models: Preliminary models were computed in WEKA explorer software version 3.6.1060 with three different algorithms: Sequential minimal optimization (SMO), logistic regression and random forest. The models were built using a preliminary version of the training sets: BitterDB as positive set and non-bitter flavors set with 2,000 diverse selected molecules from ChEBI as negative set. Class imbalance was addressed by oversampling the positive set.Ensemble methods models: Fitensemble and TreeBagger algorithms from Matlab Machine Learning (ML) toolbox (version R2015a; Mathworks, Inc., MA, USA) were used to train different preliminary models.The AdaBoost classifier was calculated with decision trees as weak learners as implemented in the fitensemble algorithm in Matlab. In order to avoid overfitting and improve performance several parameters were adjusted by examining their influence on the mean squared error (MSE) in the train set with respect to the MSE of the test set: i) Learning rate: was set to 0.15 (default is 1.0) ii) Number of trees: was set to 200 iii) Maximum number of splits in each tree (‘maxNumSplits’), and minimum number of samples in tree leaf (‘MinLeafSize’), were set to 15 (default is 5) and 5 (default is 1) respectively. The last two parameters restricted each decision tree depth.The AdaBoost algorithm is calculated in several iterations, in each iteration a new decision tree is calculated and a weight vector is adjusted to learn from previous misclassified samples. Weights of the misclassified samples are increased while weights of the correctly classified samples are reduced. In each iteration the algorithm minimizes the weighted classification error (see equation 1, equation was adapted from61).Equation 1: Weighted classification error61

$$\varepsilon t=\sum _{n=1}^{N}{d}_{n}^{(t)}I({y}_{n}\ne {h}_{t}({x}_{n}))$$

(1)

n is the descriptors vector for sample n,

n is the true label for sample n,

is the prediction of the learner with index t,I is the indicator function.

${d}_{n}^{(t)}$ is the weight of sample n at step t

Prediction with AdaBoostThe predictions are made by calculating the weighted average of the predictions given by each tree in the ensemble. The weights for each tree are determined according to each tree weighted classification error, such that predictions from trees with lower weighted classification error get higher weight (see equation 2, the equation is adapted from61).$$f(x)=\sum _{t=1}{\alpha }_{t}{h}_{t}(x)$$

(2)

where:$${\alpha }_{t}=\frac{1}{2}log\frac{1-{\varepsilon }_{t}}{{\varepsilon }_{t}}$$

X new data

is the prediction of the learner with index t,Imbalanced data and within-class imbalanceTwo main issues need to be considered while designing the prediction model: Imbalanced and Within-class imbalance62.Imbalanced data - The negative set is almost three times larger than the positive set. Machine learning approaches that aim to minimize the number of mistakes would result in overall high accuracy (accuracy = percentage of the correctly classified samples out of the total samples), even though the sensitivity (sensitivity = percentage of the correctly classified positive samples out of the total positive samples) will be low62. In our case, this would lead to predicting most of the samples as “non-bitter”. Our goal is to avoid this situation.Within-class imbalance62 - The negative set is comprised of three subsets; non-bitter flavors, tasteless and sweet. The number of compounds in the tasteless and sweet subsets is much smaller than in the non-bitter flavors set, but they occupy a different chemical space.Both issues are addressed in the final model by setting the weights in the initial observation weight vector. The observation weight vector length is the number of compounds; each compound gets a proportional weight. Weight vector was calculated according to compounds set size such that the sum of all bitter samples weight is 0.5 and the sum of all non-bitter samples weight is 0.5. The sum of weights in each non-bitter subset was set to: sweet: 0.16 tasteless 0.16 and flavors 0.18.Model performanceDatasets were randomly divided to train and hold out test set containing 70% and 30% of the molecules respectively, ensuring the original proportions of the different subsets (bitter, flavors, sweet, and tasteless). The different models were trained only on the train set, and the four parameters described above (learning rate, number of trees, ‘maxNumSplits’, MinLeafSize’) were adjusted by their performance in the training set and with respect to their performance in the hold out test set to avoid overfitting. The external validation sets were used only to assess the model.Classification models were evaluated by means of sensitivity (Se) and specificity (Sp) of classes. Se describes the true positive rate i.e how many positive samples were correctly identified as positive. Sp describes the true negative rate i.e how many negative samples were identified as negative.$$Sp=\frac{TN}{TN+FP\,}\,Se=\frac{TP}{TP+FN}$$

(3)

where TP, TN, FP and FN represent the number of true positives, true negatives, false positives, and false negatives, respectively.Descriptor’s contribution to the model’s performance was calculated using the Predictor Importance procedure in the fitensemble function is Matlab (version R2015a; Mathworks, Inc., MA, USA)63. This procedure is summing the changes in the MSE in each tree between the original split MSE in the parent node and the total MSE for the two children.Sensory experiment for validation of predictionsThe experiment included 12 participants (8 females, 4 males, mean age = 30.41, age range: 24–42), with no reported pregnancy, food intolerance, allergies or use of medication. The participants were instructed not to consume anything other than water for 1 h prior to the experiment. During the experiment the participants used nose clips. Using Compusense Cloud on-line software (Compusense Inc., Guelph, ON, Canada) the participants were guided to swish each one of the eight solutions (6 solutions of predicted non-bitter compounds, water and a known bitter compound quinine) for 5 sec. without swallowing. Between sample tastings, participants were instructed to rinse their mouth with water and wait 30 seconds. The order of the solutions was randomly assigned to each participant by the Compusense software. After sipping and spitting of each solution, participants had to evaluate its bitterness and sweetness by using 9-Likert Scale on Compusense Cloud. 9-Likert scale ranged from 1 (no sensation) to 9 (extremely strong sensation). In addition, participants had to report the dominant taste of each solution and any additional tastes they recognized. All research procedures were performed according to relevant guidelines and regulations and were ethically approved by the “Committee for the Use of Human Subjects in Research in The Robert H. Smith Faculty of Agriculture Food and environment, the Hebrew University of Jerusalem”, informed consent was obtained from all participants.ReagentsEthanethiol, 2-Decanone, Diethyl disulfide, 2,5-Dimethylthiophene, trans-2-Pentenoic acid, cis-3-Hexenyl acetate and quinine sulfate were purchased from Sigma (CAS Numbers: 75-08-1, 693-54-9, 110-81-6, 638-02-8, 13991-37-2, 3681-71-8 and 207671-44-1 respectively). All compounds were dissolved to a final concentration of 0.5 mM in double distilled water (Millipore-filtered). The solutions concentration was selected to be 0.5 mM, higher than reported detection thresholds of known bitter compounds, (such as PROP and quinine64), and safe.Statistical analysesStatistical tests were conducted using JMP Pro 13 (SAS). Data were first analyzed using ANOVA with participants as a random effect. Thereafter the Dunnett test was used to compare mean bitterness for each solution to the water stimuli. Significance was set at p < 0.05, and two-tailed tests were used where relevant.Data AvailabilityTraining and test sets generated and/or analysed in the current study are available from the corresponding author upon reasonable request.Matlab code for BitterPredict is provided via BitterDB http://bitterdb.agri.huji.ac.il/dbbitter.php#BitterPredict and via GitHub repository https://github.com/Niv-Lab/BitterPredict1. The allowed input is CSV or Excel file holding QikProp and physicochemical descriptors calculated with Schrödinger as listed in detail via Supplementary Table S3.The external sets used for validation are included in the Supplementary file validation.xls online.The prospective predictions on DrugBank, FooDB, ChEBI and ZINC natural products are included in Supplementary file prospective_prediction_sets.xls online and via Ambinter (http://www.ambinter.com) software website in the following links:FooDB: http://www.ambinter.com/moleditor/web/display/c6ff1327

DrugBank: http://www.ambinter.com/moleditor/web/display/5fb411e8

ChEBI: http://www.ambinter.com/moleditor/web/display/6dbe122e

Natural Products: http://www.ambinter.com/moleditor/web/display/6bcc1278

Predicted molecules can be browsed and filtered by several descriptors. Users can download the data in different formats and search for similar compounds.Summary of the supplementary data files can be found in Supplementary Table S1.Requests for predictions for additional datasets can be submitted to the authors via email request.

ReferencesChandrashekar, J., Hoon, M. A., Ryba, N. J. P. & Zuker, C. S. The receptors and cells for mammalian taste. Nature

444, 288–294 (2006).Article

ADS

CAS

PubMed

Google Scholar

Meyerhof, W. et al. The molecular receptive ranges of human TAS2R bitter taste receptors. Chem Senses

35, 157–170, https://doi.org/10.1093/chemse/bjp092 (2010).Article

CAS

PubMed

Google Scholar

Wiener, A., Shudler, M., Levit, A. & Niv, M. Y. BitterDB: a database of bitter compounds. Nucleic Acids Research

40, D413–D419, https://doi.org/10.1093/nar/gkr755 (2012).Article

CAS

PubMed

Google Scholar

Tizzano, M. & Finger, T. E. Chemosensors in the Nose: Guardians of the Airways. Physiology

28, 51–60, https://doi.org/10.1152/physiol.00035.2012 (2013).Article

CAS

PubMed

PubMed Central

Google Scholar

Finger, T. E. & Kinnamon, S. C. Taste isn’t just for taste buds anymore. F1000 Biology Reports

3, 20, https://doi.org/10.3410/B3-20 (2011).Article

PubMed

PubMed Central

Google Scholar

Shah, A. S., Ben-Shahar, Y., Moninger, T. O., Kline, J. N. & Welsh, M. J. Motile Cilia of Human Airway Epithelia Are Chemosensory. Science

325, 1131–1134, https://doi.org/10.1126/science.1173869 (2009).Article

ADS

CAS

PubMed

PubMed Central

Google Scholar

Behrens, M. & Meyerhof, W. Gustatory and extragustatory functions of mammalian taste receptors. Physiology & Behavior

105, 4–13 (2011).Article

CAS

Google Scholar

Lee, R. J. et al. T2R38 taste receptor polymorphisms underlie susceptibility to upper respiratory infection. The Journal of Clinical Investigation

122, 4145–4159, https://doi.org/10.1172/JCI64240 (2012).Article

CAS

PubMed

PubMed Central

Google Scholar

Liggett, S. B. Bitter taste receptors on airway smooth muscle as targets for novel bronchodilators. Expert Opinion on Therapeutic Targets

17, 721–731, https://doi.org/10.1517/14728222.2013.782395 (2013).Article

CAS

PubMed

PubMed Central

Google Scholar

Gerthoffer, W. T., Solway, J. & Camoretti-Mercado, B. Emerging targets for novel therapy of asthma. Current opinion in pharmacology

13, 324–330, https://doi.org/10.1016/j.coph.2013.04.002 (2013).Article

CAS

PubMed

PubMed Central

Google Scholar

Douglas, J. E., Saunders, C. J., Reed, D. R. & Cohen, N. A. A role for airway taste receptor modulation in the treatment of upper respiratory infections. Expert Review of Respiratory Medicine

10, 157–170, https://doi.org/10.1586/17476348.2016.1135742 (2016).Article

CAS

PubMed

PubMed Central

Google Scholar

Bahia, M. S., Nissim, I. & Niv, M. Bitterness prediction in-silico: A step towards better drugs. International Journal of Pharmaceutics (2017).Mennella, J. A. & Beauchamp, G. K. Optimizing Oral Medications for Children. Clinical therapeutics

30, 2120–2132, https://doi.org/10.1016/j.clinthera.2008.11.018 (2008).Article

PubMed

PubMed Central

Google Scholar

Mennella, J. A., Spector, A. C., Reed, D. R. & Coldwell, S. E. The Bad Taste of Medicines: Overview of Basic Research on Bitter Taste. Clinical therapeutics

35, 1225–1246, https://doi.org/10.1016/j.clinthera.2013.06.007 (2013).Article

PubMed

PubMed Central

Google Scholar

Drewnowski, A. & Gomez-Carneros, C. Bitter taste, phytonutrients, and the consumer: a review. The American Journal of Clinical Nutrition

72, 1424–1435 (2000).CAS

PubMed

Google Scholar

Rodgers, S., Glen, R. C. & Bender, A. Characterizing bitterness: identification of key structural features and development of a classification model. J Chem Inf Model

46, 569–576, https://doi.org/10.1021/ci0504418 (2006).Article

CAS

PubMed

Google Scholar

Eric Walters, D. How are bitter and sweet tastes related? Trends in Food Science & Technology

7, 399–403 (1996).Article

Google Scholar

Takahashi, Y., Miyashita, Y., Tanaka, Y., Abe, H. & Sasaki, S. A consideration for structure-taste correlations of perillartines using pattern-recognition techniques. Journal of Medicinal Chemistry

25, 1245–1248, https://doi.org/10.1021/jm00352a030 (1982).Article

CAS

PubMed

Google Scholar

Cravotto, G., Nano, G. M., Binello, A., Spagliardi, P. & Seu, G. Chemical and biological modification of cynaropicrin and grosheimin: a structure–bitterness relationship study. Journal of the Science of Food and Agriculture

85, 1757–1764, https://doi.org/10.1002/jsfa.2180 (2005).Article

CAS

Google Scholar

Scotti, L. et al. Quantitative elucidation of the structure–bitterness relationship of cynaropicrin and grosheimin derivatives. Food Chemistry

105, 77–83 (2007).Article

CAS

Google Scholar

Spillane, W. J., Feeney, B. G. & Coyle, C. M. Further studies on the synthesis and tastes of monosubstituted benzenesulfamates. A semi-quantitative structure–taste relationship for the meta-compounds. Food Chemistry

79, 15–22 (2002).Article

CAS

Google Scholar

Soltani, S. et al. QSBR Study of Bitter Taste of Peptides: Application of GA-PLS in Combination with MLR, SVM, and ANN Approaches. BioMed Research International

2013, 13, https://doi.org/10.1155/2013/501310 (2013).Article

Google Scholar

Ney, K. H. In Food Taste Chemistry Vol. 115 ACS Symposium Series Ch. 6, 149-173 (American Chemical Society, 1979).Levit, A. et al. The bitter pill: clinical drugs that activate the human bitter taste receptor TAS2R14. The FASEB Journal

28, 1181–1197, https://doi.org/10.1096/fj.13-242594 (2014).Article

CAS

PubMed

Google Scholar

Roland, W. S. U. et al. Bitter Taste Receptor Activation by Flavonoids and Isoflavonoids: Modeled Structural Requirements for Activation of hTAS2R14 and hTAS2R39. Journal of Agricultural and Food Chemistry

61, 10454–10466, https://doi.org/10.1021/jf403387p (2013).Article

CAS

PubMed

Google Scholar

Karaman, R. et al. Probing the Binding Pocket of the Broadly Tuned Human Bitter Taste Receptor TAS2R14 by Chemical Modification of Cognate Agonists. Chemical Biology & Drug Design

88, 66–75, https://doi.org/10.1111/cbdd.12734 (2016).Article

CAS

Google Scholar

Bender, A., Mussa, H. Y., Glen, R. C. & Reiling, S. Molecular Similarity Searching Using Atom Environments, Information-Based Feature Selection, and a Naïve Bayesian Classifier. Journal of Chemical Information and Computer Sciences

44, 170–178, https://doi.org/10.1021/ci034207y (2004).Article

CAS

PubMed

Google Scholar

Huang, W. et al. BitterX: a tool for understanding bitter taste in humans. Scientific Reports

6, 23450, https://doi.org/10.1038/srep23450 (2016).Article

ADS

CAS

PubMed

PubMed Central

Google Scholar

Hastings, J. et al. The ChEBI reference database and ontology for biologically relevant chemistry: enhancements for 2013. Nucleic Acids Research

41, D456–D463, https://doi.org/10.1093/nar/gks1146 (2012).Article

PubMed

PubMed Central

Google Scholar

Rojas, C. et al. Quantitative structure–activity relationships to predict sweet and non-sweet tastes. Theoretical Chemistry Accounts

135, 1–13, https://doi.org/10.1007/s00214-016-1812-1 (2016).Article

CAS

Google Scholar

Sahigara, F. et al. Comparison of Different Approaches to Define the Applicability Domain of QSAR Models. Molecules

17, 4791 (2012).Article

CAS

PubMed

Google Scholar

Burdock, G. A. Fenaroli’s handbook of flavor ingredients. 5 edn, (CRC Press, 2005).In Hazardous Substances Data Bank [Internet].: National Library of Medicine (US); Bethesda (MD) Available from: http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen?HSDB (2011).Arnoldi, A., Bassoli, A., Borgonovo, G., Merlini, L. & Morini, G. Synthesis and Structure−Activity Relationships of Sweet 2-Benzoylbenzoic Acid Derivatives. Journal of Agricultural and Food Chemistry

45, 2047–2054, https://doi.org/10.1021/jf960755i (1997).Article

CAS

Google Scholar

Ley, J. P., Blings, M., Paetz, S., Krammer, G. E. & Bertram, H.-J. New Bitter-Masking Compounds: Hydroxylated Benzoic Acid Amides of Aromatic Amines as Structural Analogues of Homoeriodictyol. Journal of Agricultural and Food Chemistry

54, 8574–8579, https://doi.org/10.1021/jf0617061 (2006).Article

CAS

PubMed

Google Scholar

Glendinning, J. I. Is the bitter rejection response always adaptive? Physiology & Behavior

56, 1217–1227 (1994).Article

CAS

Google Scholar

Freund, Y. & Schapire, R. E. In European conference on computational learning theory 23–37 (Springer, 1995).Tropsha, A. Best Practices for QSAR Model Development, Validation, and Exploitation. Molecular Informatics

29, 476–488, https://doi.org/10.1002/minf.201000061 (2010).Article

CAS

PubMed

Google Scholar

Mancuso, G., Borgonovo, G., Scaglioni, L. & Bassoli, A. Phytochemicals from Ruta graveolens Activate TAS2R Bitter Taste Receptors and TRP Channels Involved in Gustation and Nociception. Molecules

20, https://doi.org/10.3390/molecules201018907 (2015).Kohl, S., Behrens, M., Dunkel, A., Hofmann, T. & Meyerhof, W. Amino Acids and Peptides Activate at Least Five Members of the Human Bitter Taste Receptor Family. Journal of Agricultural and Food Chemistry

61, 53–60, https://doi.org/10.1021/jf303146h (2013).Article

CAS

PubMed

Google Scholar

Pickrahn, S., Sebald, K. & Hofmann, T. Application of 2D-HPLC/Taste Dilution Analysis on Taste Compounds in Aniseed (Pimpinella anisum L.). Journal of Agricultural and Food Chemistry

62, 9239–9245 (2014).Article

CAS

PubMed

Google Scholar

Tan, Q., Zhang, L., Zhang, L., Teng, Y. & Zhang, J. Design and Evaluation of an Economic Taste-Masked Dispersible Tablet of Pyridostigmine Bromide, a Highly Soluble Drug with an Extremely Bitter Taste. Chemical and Pharmaceutical Bulletin

60, 1514–1521, https://doi.org/10.1248/cpb.c12-00629 (2012).Article

CAS

PubMed

Google Scholar

Soares, S. et al. Different Phenolic Compounds Activate Distinct Human Bitter Taste Receptors. Journal of Agricultural and Food Chemistry

61, 1525–1533, https://doi.org/10.1021/jf304198k (2013).Article

CAS

PubMed

Google Scholar

AVERSION TECHNOLOGIES, INC - colocynth natural bitterant http://aversiontech.com/products/naturals/colocynth/ (2015).Baxter, H., Harborne, J. B. & Moss, G. P. Phytochemical dictionary: a handbook of bioactive compounds from plants. (CRC press, 1998).Wishart, D. S. et al. DrugBank: a comprehensive resource for in silico drug discovery and exploration. Nucleic Acids Research

34, D668–D672, https://doi.org/10.1093/nar/gkj067 (2006).Article

CAS

PubMed

Google Scholar

Kim, S. et al. PubChem Substance and Compound databases. Nucleic Acids Research

44, D1202–D1213, https://doi.org/10.1093/nar/gkv951 (2016).Article

CAS

PubMed

Google Scholar

Pence, H. E. & Williams, A. ChemSpider: An Online Chemical Information Resource. Journal of Chemical Education

87, 1123–1124, https://doi.org/10.1021/ed100697w (2010).Article

ADS

CAS

Google Scholar

Foster, S. R. et al. Expression, Regulation and Putative Nutrient-Sensing Function of Taste GPCRs in the Heart. PLoS ONE

8, e64579, https://doi.org/10.1371/journal.pone.0064579 (2013).Article

ADS

CAS

PubMed

PubMed Central

Google Scholar

Sterling, T. & Irwin, J. J. ZINC 15 – Ligand Discovery for Everyone. Journal of Chemical Information and Modeling

55, 2324–2337, https://doi.org/10.1021/acs.jcim.5b00559 (2015).Article

CAS

PubMed

PubMed Central

Google Scholar

War, A. R. et al. Mechanisms of plant defense against insect herbivores. Plant Signaling & Behavior

7, 1306–1320, https://doi.org/10.4161/psb.21663 (2012).Article

Google Scholar

Hanson, J. R. In Natural Products: The Secondary Metabolites Vol. 17 (ed James R. Hanson) 1–34 (The Royal Society of Chemistry, 2003).Li, D. & Zhang, J. Diet Shapes the Evolution of the Vertebrate Bitter Taste Receptor Gene Repertoire. Molecular Biology and Evolution

31, 303–309, https://doi.org/10.1093/molbev/mst219 (2014).Article

CAS

PubMed

Google Scholar

Ling-Ling, H. U. & Smallest, P. S. bitter taste receptor (T2Rs) gene repertoire in carnivores. Zoological Research

34, 24575-E, https://doi.org/10.11813/j.issn.0254-5853.2013.E3.E75 (2013).

Google Scholar

Belitz, H. D. & Wieser, H. Bitter compounds: Occurrence and structure‐activity relationships. Food Reviews International

1, 271–354, https://doi.org/10.1080/87559128509540773 (1985).Article

CAS

Google Scholar

Guigoz, Y. S. & Bitter, J. Peptides, Occurrence And Structure. Chemical Senses

2, 71–84, https://doi.org/10.1093/chemse/2.1.71 (1976).Article

CAS

Google Scholar

Cho, M. J., Unklesbay, N., Hsieh, F.-h & Clarke, A. D. Hydrophobicity of Bitter Peptides from Soy Protein Hydrolysates. Journal of Agricultural and Food Chemistry

52, 5895–5901, https://doi.org/10.1021/jf0495035 (2004).Article

CAS

PubMed

Google Scholar

Ye, H. et al. HIT: linking herbal active ingredients to targets. Nucleic Acids Research

39, D1055–D1059, https://doi.org/10.1093/nar/gkq1165 (2011).Article

CAS

PubMed

Google Scholar

Kang, H. et al. HIM-herbal ingredients in-vivo metabolism database. Journal of Cheminformatics

5, 28, https://doi.org/10.1186/1758-2946-5-28 (2013).Article

CAS

PubMed

PubMed Central

Google Scholar

Hall, M. et al. The WEKA data mining software: an update. ACM SIGKDD explorations newsletter

11, 10–18 (2009).Article

Google Scholar

Ensemble Methods - MATLAB, AdaBoostM1, https://www.mathworks.com/help/stats/ensemble-methods.html#bsw8aue (2017).Sun, Y., Wong, A. K. C. & Kamel, M. S. Classification of imbalanced data: a review. International Journal of Pattern Recognition and Artificial Intelligence

23, 687–719, https://doi.org/10.1142/S0218001409007326 (2009).Article

Google Scholar

Estimates of predictor importance - MATLAB, https://www.mathworks.com/help/stats/compactregressionensemble.predictorimportance.html (2017).Dubovski, N., Ert, E. & Niv, M. Y. Bitter mouth-rinse affects emotions. Food Quality and Preference

60, 154–164, https://doi.org/10.1016/j.foodqual.2017.04.007 (2017).Article

Google Scholar

Download referencesAcknowledgementsFunding by the ISF 494/16 and ISF-NSFC 2463/16 grants to MYN, and the Robert H. Smith PHD scholarship to ADW are gratefully acknowledged. The authors participate in the European COST Actions CM1027 (GLISTEN) and CA15135 (Mu.Ta.Lig). The authors thank Prof. Hanoch Senderowitz and his lab members for helpful discussions on machine learning methods, Dr. Einav Malach for technical assistance, and Dr. Hillary Voet for advice on statistical analysis.Author informationAuthors and AffiliationsInstitute of Biochemistry, Food Science and Nutrition, The Robert H. Smith Faculty of Agriculture, Food, and Environment, The Hebrew University of Jerusalem, Rehovot, 76100, IsraelAyana Dagan-Wiener, Ido Nissim, Natalie Ben Abu & Masha Y. NivThe Fritz Haber Center for Molecular Dynamics, The Hebrew University of Jerusalem, Jerusalem, 91904, IsraelAyana Dagan-Wiener, Ido Nissim, Natalie Ben Abu & Masha Y. NivDeFENS-Department for Food, Environmental and Nutritional Sciences, University of Milan, Via Celoria 2, Milano, 20133, ItalyGigliola Borgonovo & Angela BassoliAuthorsAyana Dagan-WienerView author publicationsYou can also search for this author in

PubMed Google ScholarIdo NissimView author publicationsYou can also search for this author in

PubMed Google ScholarNatalie Ben AbuView author publicationsYou can also search for this author in

PubMed Google ScholarGigliola BorgonovoView author publicationsYou can also search for this author in

PubMed Google ScholarAngela BassoliView author publicationsYou can also search for this author in

PubMed Google ScholarMasha Y. NivView author publicationsYou can also search for this author in

PubMed Google ScholarContributionsA.D.W. and M.Y.N. designed the study and wrote the paper, G.B. and A.B. established the UNIMI set, A.D.W. and I.N. prepared the datasets and carried out literature searches. N.B.A. conducted sensory tests. A.D.W. constructed and optimized the classifier and analyzed the results. All authors reviewed and approved the final manuscript.Corresponding authorCorrespondence to

Masha Y. Niv.Ethics declarations

Competing Interests

The authors declare that they have no competing interests.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Electronic supplementary materialSupplementary dataSupplementary Dataset 1Supplementary Dataset 2Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissionsAbout this articleCite this articleDagan-Wiener, A., Nissim, I., Ben Abu, N. et al. Bitter or not? BitterPredict, a tool for predicting taste from chemical structure.

Sci Rep 7, 12074 (2017). https://doi.org/10.1038/s41598-017-12359-7Download citationReceived: 07 April 2017Accepted: 07 September 2017Published: 21 September 2017DOI: https://doi.org/10.1038/s41598-017-12359-7Share this articleAnyone you share the following link with will be able to read this content:Get shareable linkSorry, a shareable link is not currently available for this article.Copy to clipboard

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Morus alba L. Leaves (WML) Modulate Sweet (TAS1R) and Bitter (TAS2R) Taste in the Studies on Human Receptors – A New Perspective on the Utilization of White Mulberry Leaves in Food Production?

Monika PrzeorMaria Jokiel

Plant Foods for Human Nutrition (2023)

RETRACTED ARTICLE: Synthesis, Pharmacological Evaluation, and Molecular Modeling Studies of New Isatin Hybrids as Potential Anticancer Agents

Rajapandi RajuKumarappan ChidambaramTapan Kumar Maity

Pharmaceutical Chemistry Journal (2023)

Consumers of Extra Virgin Olive Oil Perceive Sensory Attributes in the Same Way as Trained Panelists?

Amanda Neris dos SantosMatheus dos Santos e SilvaCamila Argenta Fante

Food Analytical Methods (2023)

BitterMatch: recommendation systems for matching molecules with bitter taste receptors

Eitan MargulisYuli SlavutskyMasha Y. Niv

Journal of Cheminformatics (2022)

Toward a general and interpretable umami taste predictor using a multi-objective machine learning approach

Lorenzo PallanteAigli KorfiatiMarco A. Deriu

Scientific Reports (2022)

CommentsBy submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Download PDF

Explore content

Research articles

News & Comment

Collections

Subjects

RSS feed

About the journal

Open Access Fees and Funding

About Scientific Reports

Contact

Journal policies

Calls for Papers

Guide to referees

Editor's Choice

Journal highlights

Publish with us

For authors

Language editing services

Submit manuscript

Search articles by subject, keyword or author

Show results from

All journals

This journal

Advanced search

Quick links

Explore articles by subject

Find a job

Guide to authors

Editorial policies

Scientific Reports (Sci Rep)

ISSN 2045-2322 (online)

nature.com sitemap

About Nature Portfolio

About us

Press releases

Press office

Discover content

Journals A-Z

Articles by subject

Protocol Exchange

Nature Index

Publishing policies

Nature portfolio policies

Open access

Author & Researcher services

Reprints & permissions

Research data

Language editing

Scientific editing

Nature Masterclasses

Research Solutions

Libraries & institutions

Librarian service & tools

Librarian portal

Open research

Recommend to library

Advertising & partnerships

Advertising

Partnerships & Services

Media kits

Branded

content

Professional development

Nature Careers

Nature

Conferences

Regional websites

Nature Africa

Nature China

Nature India

Nature Italy

Nature Japan

Nature Korea

Nature Middle East

Privacy

Policy

Use

of cookies

Your privacy choices/Manage cookies

Legal

notice

Accessibility

statement

Terms & Conditions

Your US state privacy rights

Close banner

Email address

I agree my information will be processed in accordance with the Nature and Springer Nature Limited Privacy Policy.

Close banner

Get the most important science stories of the day, free in your inbox.

13 Healthiest Bitter Foods - Bitter Fruits and Vegetables for Digestive Health

ealthiest Bitter Foods - Bitter Fruits and Vegetables for Digestive HealthSearchSubscribeMy BookmarksGH+ Member PortalProduct ReviewsHome IdeasFood + RecipesAll RecipesHealthBeauty + StyleHolidaysLifeAbout UsAwardsEventsNewsletterFollowShopPromotionsOther EditionsPrivacy NoticeTerms Of UseSkip to ContentProduct ReviewsLifeHealthFoodBeautySubscribesign inBest Walking ShoesBest LuggageBest Skincare RoutineBest Mattresses of 2024Bathing Suits for WomenHealthDiet & NutritionThese 13 Bitter Foods Can Revolutionize Your Gut Health, According to NutritionistsThese 13 Bitter Foods Can Revolutionize Your Gut Health, According to NutritionistsThere's a science behind why these fruits and vegetables can trigger optimal digestion in time, making these staples worth your consideration.By Zee KrsticPublished: Feb 8, 2022Save ArticleVICUSCHKABitter foods can be hard to swallow (literally!) for some, as their strong flavors can overpower sweet, salty or umami notes of an otherwise tasty dish. But those who make it their mission to skillfully incorporate more bitter flavors into their diet are enjoying a rich payoff when it comes to gut health, says Amy Fischer, MS, RD, CDN, a registered dietitian within the Good Housekeeping Institute.The surprising truth about bitter foods is that they contain plant-based chemicals that can streamline how your body reacts to nutritious meals, an added bonus to the fact that most bitter foods are entirely nutritious on their own. "Bitters — and bitter-tasting herbs and foods — have been used for millennia as a digestive aid," Fischer adds. "In a nutshell, bitter foods increase saliva production and start the digestion process because of their bitter flavor."Read More:Foods to Fight ConstipationWhich kinds of foods can you consider bitter, you may wonder? They may already be grocery staples you know and love — think things like a daily cup of coffee, fresh cranberries, crunchy kale in your favorite salad. Foods and beverages like these contain bitter elements that stimulate your mouth's taste buds, which in turn activate saliva production when you're eating. From there, Fischer says that excess saliva triggers gastric acid to aid in immediate digestion, later stimulating bile flow in your gut. "Stomach acid is a good thing, and you need enough to help digest food that you eat. Therefore, incorporating bitter foods into your diet can help with bowel regularity and constipation, because these foods help to keep food moving through the digestive tract," Fischer adds.If you're noticing that you're having irregularity, or are experiencing issues with bloating or constipation frequently, slowly ramping up how frequently you eat these bitter nutrition powerhouses below may help you achieve normalcy when it comes to digestion. Remember: Bitter foods won't fix an irregular gut overnight, no matter how much you eat, so be sure to slowly incorporate the following into your diet to avoid gas or acid reflux, Fischer says. Looking to streamline digestive health with the help of healthy bitter flavors? All of the following ingredients can be incorporated into your favorite recipes, aiding digestion while bringing their unique nutritional benefits into the mix as well.Cavan Images//Getty Images1) Kale Maybe you're missing it beneath a generous drizzle of delicious salad dressing, but kale does qualify as a bitter flavor profile that promotes digestion — alongside the fact that it's a fiber powerhouse, which helps you stay regular. Kale is loaded with nutrients and antioxidants, particularly vitamins A and K as well as calcium and potassium. It contains plant-based glucosinolates, a group of sulfur-containing compounds that help to better regulate your liver health, Fischer says. "Kale contains prebiotics that promote good gut health, as it helps to increase the amount of good gut bacteria, helping with digestion," she adds. "Kale works to reduce bad cholesterol, known as LDL, by binding to extra circulating cholesterol in your system and carrying it out through elimination." 2) ArugulaRead More:Why Do I Always Feel So Bloated?A close relative to kale, peppery arugula is chock full of vitamins, minerals and antioxidants, and also works to add more fiber into a diet. In similarity to Brussels sprouts and broccoli rabe, arugula's bitter flavor profile comes from glucosinolates, which is responsible for a suite of vegetables' bitter taste. While glucosinolates may deter animals and pests from munching on plants in the wild — "Glucosinolates provide plants with protection against bugs, as they act as a natural pesticide," Fischer explains — you shouldn't be wary of arugula yourself. The strong, bitter flavors from these compounds kick your taste buds into gear, which helps promote digestion.3) Broccoli RabeA cruciferous pick that's part of the brassica family, Fischer points to broccoli rabe as one of the richest sources of vitamins A, C and K on this list. It may also equally be one of the most bitter and offensive to sensitive palates, which is why many of the tastiest broccoli rabe recipes call for lemon or citrus to break away from bitter flavors. "Lemon can also aid in the absorption of iron in broccoli rabe," she adds. "Vitamins A&K are fat-soluble, so you'll want to enjoy this with a generous serving of healthy fats."Claudia Totir4) RadicchioAlongside other produce in the chicory family (think Belgian endives, escarole and other fall favorites), radicchio carries many nutritious qualities despite its overtly bitter flavor. Home cooks often use radicchio to add an earthy touch to protein-heavy dishes and sautes, or as a salad base to pair nicely with fresh citrus and other zesty seasonings. "Radicchio is rich in fiber, zinc and the fat-soluble vitamin K, so you'll want to enjoy it with a healthy fat, like olive oil," Fischer says. "Fiber also helps to keep blood sugar more stable and it keeps you feeling fuller longer, which is helpful with weight management."5) Brussels SproutsAnother vegetable loaded with glucosinolates, Brussels sprouts hold a significant amount of naturally occurring potassium in each bite, alongside Vitamins B and C. Its reputation may be worse among those who are averse to its strong flavor profile, but there's a reason why parents everywhere are still harping on Brussels sprouts — their glucosinolates composition (and many others on this list) may purportedly work to lower the risk of cancer over time, though research remains divided on how or why. More research needs to be done on the cancer front, but Fischer and nutrition experts everywhere remain certain this fiber-packed vegetable can do wonders for your gut over time. 6) Dandelion Greens "A lot of people think of dandelion as a pesky weed, but it has beneficial properties — the leaves are bitter and contain inulin, which can help lower bad cholesterol and may help to keep blood sugar more stable," explains Fischer. "Dandelion contains vitamins A, C, K, as well as folate, iron, calcium and potassium, which is a natural diuretic," she adds.Usually, dandelion greens are used in salads, spun into green juice or even processed into Dandelion tea to help with digestive issues.VICUSCHKA7) EndivesAnother member of the chicory family, Fischer says endives are set apart due to their inulin composition, an indigestible prebiotic fiber that occurs naturally within this crisp leaf vegetable. "Inulin can help with holistic digestion as it promotes good bacteria," she adds. Endives also are rich in vitamin A, C and E, packing an extra punch of electrolytes found within their potassium makeup.8) GrapefruitCitric acid is what makes a fresh grapefruit feel intensely bitter in your mouth — and when consumed in excess, can contribute to chronic irritation for those suffering from irritable bowel syndrome or heartburn. While other kinds of fresh citrus have been known to aid digestion — oranges, lemons, and limes specifically — grapefruits, in particular, have a high fiber composition and are loaded with water, Fischer says, aiding in feeling properly satiated until the next meal (hydration is key!). "It's also a rich source of Vitamin C, which can promote collagen production; key for healthy skin, hair and nails."9) CacaoChocolate isn't the same as cacao, which is a component of the final product that is entirely bitter in its pure, unadulterated form — and much more redeeming for your holistic health, in addition to a digestive aid. Unsweetened cacao is chock full of magnesium, potassium, iron and zinc, as well as flavonoids, which is a form of a polyphenol that aids your body in fighting inflammation over time. In the healthiest chocolate bars, significant magnesium content has been tied to relaxation and sleep benefits as well. "Cacao is also full of antioxidants, a great source of electrolytes and minerals like magnesium, contains iron, and potassium, which is a natural diuretic," Fischer adds.Gina Pricope / Getty Images10) CranberriesNot to be mistaken with processed cranberry juice, which can contain lots of added sugar, this bitter fruit is likely the most popular on this list. Organic cranberries are extremely tart and get the digestion process 'moving' sooner rather than later. They can be added into fresh salads, marinades, or even smoothies; and regularly incorporating them into your diet may aid digestive discomfort at large."Studies have shown that regular consumption of sugar-free, tart cranberry juice could help to suppress H-pylori infection, a common stomach infection," Fischer says. "And cranberry juice may help prevent stomach ulcers from developing, or at least treat symptoms in people that already have ulcers." 11) CoffeeIf you feel like you can't get your day started properly without a cup of coffee, you wouldn't be the only one — coffee is bitter and tart, promoting saliva production and jumpstarting the digestion process (key at breakfast!). And research suggests that coffee may be the single greatest contributor to antioxidant intake in your diet, as it contains many chlorogenic acids, which work to prevent vision loss or blindness as you age. Plus, antioxidants can help to fight inflammation across your body. Stick to caffeinated coffee, though, as decaf coffee loses many of its antioxidants in this process, and new research suggests unfiltered coffee may lead to increased cholesterol consumption overall.12) Apple Cider VinegarThere isn't a lot confirmed about apple cider vinegar, as research is divided on how this tart, bitter oily mix actually boosts health in a mechanical way. But it's flavor profile can aid in digestion, which is why people take vinegar shots. "There's a lot of mixed data on the health benefits of apple cider vinegar but it is acidic and sour and can aid in stimulating the salivary glands and begin the digestion process," Fischer added.13) Herbal PeppermintPeppermint oil itself has been linked in research to irritable bowel syndrome (IBS), as a form of alternative treatment to alleviate digestive symptoms. The evidence, published in 2014, specifically suggested that oil (when used appropriately) can relax your digestive tract by relieving muscle contractions or spams at large. At large, evidence seems to suggest it's a great supplement to discuss with your doctor. "A review of nine research studies found that peppermint oil is 'safe and effective short-term treatment for IBS,'" Fischer says.Read More:The Best Multivitamins for WomenZee KrsticContent Strategy Manager, Hearst MagazinesZee Krstic is a content strategy manager for Hearst Magazines, focusing on SEO optimization and other editorial strategies for four brands, including Country Living, House Beautiful, ELLE Decor and VERANDA. He previously served as Health Editor for Good Housekeeping between 2019 and 2023, covering health news, diet and fitness trends as well as executing wellness product reviews in conjunction with the Good Housekeeping Institute. Prior to joining Hearst, Zee fostered a strong background in women's lifestyle media with eight plus years of editorial experience, including as a site-wide editor at Martha Stewart Living after developing a nutrition background as an assistant editor at Cooking Light. Zee produces service-based health coverage, as well as design and travel content, for Hearst brands on a contributor basis; he has written about food and dining for Time, among other publications.Advertisement - Continue Reading BelowDiet & Nutrition13 Foods That'll Help You Sleep Through the NightFrom Good Housekeeping for The Vitamin Shoppe4 Health-Focused Smoothies To Brighten Your DayWhat to Eat for Healthy HairThe Best Multivitamins for WomenAdvertisement - Continue Reading BelowGood Housekeeping Dietary Supplements MethodologyTop Health Benefits of Coconut WaterGet Cooking With Celebrity Chef Jason RobertsThe Surprising Health Benefits of Black RicePopeye: Our Newest GH Nutritionist Approved MemberCan Late-Night Eating Affect Your Heart Health?The Best Energy Drinks10 Superfoods to Add to Your Cart ASAPAdvertisement - Continue Reading BelowSubscribeGive GH as a GiftOther Hearst SubscriptionsNewsletterAbout UsContact UsWork for Good HousekeepingMedia KitAdvertise OnlineCustomer ServiceEvents & PromotionsGiveawaysA Part of Hearst Digital MediaGood Housekeeping participates in various affiliate marketing programs, which means we may get paid commissions on editorially chosen products purchased through our links to retailer sites.©2024 Hearst Magazine Media, Inc. All Rights Reserved.Privacy NoticeCA Notice at CollectionYour CA Privacy Rights/Shine the LightDAA Industry Opt OutTerms of UseSite MapCookies Choi

Page restricted | ScienceDirect

Your Browser is out of date.

Update your browser to view ScienceDirect.

View recommended browsers.

Request details:

Request ID: 863cdebce9c46e4c-HKG

IP: 49.157.13.121

UTC time: 2024-03-13T14:56:20+00:00

Browser: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/97.0.4692.71 Safari/537.36

About ScienceDirect

Shopping cart

Contact and support

Terms and conditions

Cookies are used by this site. By continuing you agree to the use of cookies.

[PDF] Traditional Chinese Bitter Flavor theory: Is there any relation with taste type II receptors? | Semantic Scholar

Skip to search formSkip to main contentSkip to account menuSemantic ScholarSemantic Scholar's LogoSearch 217,141,767 papers from all fields of scienceSearchSign InCreate Free AccountDOI:10.1016/J.EUJIM.2016.04.011Corpus ID: 75799494Traditional Chinese Bitter Flavor theory: Is there any relation with taste type II receptors?@article{Zhang2016TraditionalCB,

title={Traditional Chinese Bitter Flavor theory: Is there any relation with taste type II receptors?},

author={Yuxin Zhang and Xing Wang and Shi-feng Wang and Yan-ling Zhang and Yanjiang Qiao},

journal={European Journal of Integrative Medicine},

year={2016},

volume={8},

pages={980-990},

url={https://api.semanticscholar.org/CorpusID:75799494}

}Yuxin Zhang, Xing Wang, +2 authors Yanjiang QiaoPublished 1 December 2016MedicineEuropean Journal of Integrative MedicineView via Publishermanuscript.elsevier.comSave to LibrarySaveCreate AlertAlertCiteShare6 CitationsBackground Citations1View All6 CitationsCitation TypeHas PDFAuthorMore FiltersMore FiltersFiltersSort by RelevanceSort by Most Influenced PapersSort by Citation CountSort by RecencyToward the Identification of Extra-Oral TAS2R Agonists as Drug Agents for Muscle Relaxation Therapies via Bioinformatics-Aided Screening of Bitter Compounds in Traditional Chinese MedicineMingzhi LuoKai NiYang JinZifan YuLinhong DengMedicine, ChemistryFront. Physiol.2019TLDRA knowledge-based approach and technological platform for identification or discovery of extra-oral TAS2R agonists that can be used as novel drug agents for muscle relaxation therapies through screening and evaluation of chemical compounds used in bitter flavored TCM is presented.Expand19PDFSaveClinical Associations of Bitter Taste Perception and Bitter Taste Receptor Variants and the Potential for Personalized HealthcareZiwen MaoW. ChengZhenwei LiManye YaoKeming SunMedicine, BiologyPharmacogenomics and personalized medicine2023TLDRIt is examined how T2R polymorphisms, expression levels and bitter taste perception can lead to varying clinical associations and healthcare management can potentially be individualized through appropriately administering drugs with bitter masking to increase compliance.Expand1[PDF]Save“ECTOPIC” GUSTATIVE AND OLFACTORY RECEPTORS IN THE BRAIN – NEW TARGETS FOR NEURODEGENERATION THERAPY?Feng Ifrim-ChenMedicine, BiologyFARMACIA2019TLDRA new (ethno)pharmacological hypothesis is elaborate according to which one of the multiple mechanisms of action of the neuro-protective plant derived tastants may be mediated by the brain TRs and ORs.Expand43 ExcerptsSaveIdentification of a specific agonist of human TAS2R14 from Radix Bupleuri through virtual screening, functional evaluation and binding studiesYuxin ZhangXing Wang+11 authors Yanjiang QiaoChemistry, MedicineScientific Reports2017TLDRSaikosaponin b (SSb) was confirmed for the first time to be a specific agonist of TAS2R14 and showed the ability to inhibit IgE-induced mast cell degranulation.Expand25PDFSaveCharacterization of key bitter compounds in Idesia polycarpa var. vestita Diels fruit by sensory-guided fractionation.Xuwen XiangQingqing Yang+5 authors Jianquan KanAgricultural and Food Sciences, ChemistryFood chemistry2023SaveSanguinarine Rapidly Relaxes Rat Airway Smooth Muscle Cells Dependent on TAS2R Signaling.Mingzhi LuoPeili Yu+5 authors Linhong DengMedicineBiological & pharmaceutical bulletin2020TLDRThe rapid relaxation effect of SA at low concentration (<1 μM) on cultured ASMCs depending on TAS2R signaling is established, indicating that SA might be developed as a useful bronchodilator in asthma therapy.Expand6PDFSave32 ReferencesCitation TypeHas PDFAuthorMore FiltersMore FiltersFiltersSort by RelevanceSort by Most Influenced PapersSort by Citation CountSort by Recency[Study on relations between transient receptor potential vanilloid 1 and pungent property of traditional Chinese medicines].Xing WangYanling ZhangYun WangZhen-zhen RenHong-juan BaoY. QiaoMedicineZhongguo Zhong yao za zhi = Zhongguo zhongyao…2014TLDRThe results showed that the matching relationship between TRPV1 agonist pharmacophore model and TCM chemical components could identify the active ingredients from pungent herbs and proposed that TRPv1 is one of the potential targets for efficient pungENT herbs.Expand4SaveThree TAS2R Bitter Taste Receptors Mediate the Psychophysical Responses to Bitter Compounds of Hops (Humulus lupulus L.) and BeerD. IntelmannC. BatramC. KuhnGesa HaseleuW. MeyerhofT. HofmannBiology2009TLDRThe subjects perceived the bitterness of the investigated compounds at higher concentrations than those predicted by the results of the in vitro experiments, and differences were shown to be due, at least in part, to interactions of the bitter substances with the oral mucosa.Expand88SaveExtraoral bitter taste receptors as mediators of off‐target drug effectsAdam A. ClarkS. LiggettS. MungerChemistry, MedicineFASEB journal : official publication of the…2012TLDRIt is proposed that any drug with a bitter taste could have unintended actions in the body through stimulation of extraoral type 2 taste receptors (T2Rs), a novel hypothesis that could explain many off‐target effects of diverse pharmaceuticals.Expand115PDFSaveThe bitter pill: clinical drugs that activate the human bitter taste receptor TAS2R14A. LevitS. Nowak+4 authors M. NivChemistry, MedicineFASEB journal : official publication of the…2014TLDRDespite immense chemical diversity of known TAS2R14 ligands, novel ligands and previously unknown polypharmacology of drugs were unraveled by in vitro screening of computational predictions, which enables rational repurposing of traditional and standard drugs for bitter taste signaling modulation for therapeutic indications.Expand89SaveReceptor Agonism and Antagonism of Dietary Bitter CompoundsA. BrockhoffM. BehrensN. RoudnitzkyG. AppendinoC. AvontoW. MeyerhofBiology, ChemistryThe Journal of Neuroscience2011TLDRIt is demonstrated that mixtures of bitter compounds, because they normally occur in human foodstuff, likely elicit bitter perception in a complex and not in a merely additive manner, and that the naturally occurring bitter taste receptor antagonists have shaped some of the pharmacological properties of the receptors, such as overlapping recognition profiles and breadth of tuning.Expand98PDFSaveThe Quantitative Ideas and Methods in Assessment of Four Properties of Chinese Medicinal HerbsJialei FuJingxiang PangXiaolei ZhaoJinxiang HanMedicineCell Biochemistry and Biophysics2014TLDRIt is hypothesize that by the use of biophoton analysis system, the four properties and meridian tropism of Chinese medicinal herbs can be quantitatively expressed.Expand15SaveBitterDB: a database of bitter compoundsAyana WienerMarina ShudlerA. LevitM. NivBiologyNucleic Acids Res.2012TLDRThe aim of BitterDB is to facilitate studying the chemical features associated with bitterness, which may contribute to predicting bitterness of unknown compounds, predicting ligands for bitter receptors from different species and rational design of bitterness modulators.Expand208[PDF]SaveA bitter herbal medicine Gentiana scabra root extract stimulates glucagon-like peptide-1 secretion and regulates blood glucose in db/db mouse.Hyo-Weon SuhKi-Beom Lee+9 authors H. JangMedicineJournal of ethnopharmacology201539SaveBerberine induces GLP-1 secretion through activation of bitter taste receptor pathways.Yunli YuGang Hao+6 authors Xiaozhou WenMedicineBiochemical pharmacology201573SaveDeveloping novel and general descriptors for traditional Chinese medicine (TCM) formulas: A case study of quantitative formula–activity relationship (QFAR) model for hypertension prescriptionsLu XuDehua DengJian-hui JiangR. YuXiumei WuYu ZhaoMedicine20115Save...1234...Related PapersShowing 1 through 3 of 0 Related Papers6 Citations32 ReferencesRelated PapersStay Connected With Semantic ScholarSign UpWhat Is Semantic Scholar?Semantic Scholar is a free, AI-powered research tool for scientific literature, based at the Allen Institute for AI.Learn MoreAboutAbout UsMeet the TeamPublishersBlog (opens in a new tab)AI2 Careers (opens in a new tab)ProductProduct OverviewSemantic ReaderScholar's HubBeta ProgramRelease NotesAPIAPI OverviewAPI TutorialsAPI Documentation (opens in a new tab)API GalleryResearchPublicationsResearchersResearch CareersPrototypesResourcesHelpFAQLibrariansTutorialsContactProudly built by AI2 (opens in a new tab)Collaborators & Attributions •Terms of Service (opens in a new tab)•Privacy Policy (opens in a new tab)•API License AgreementThe Allen Institute for AI (opens in a new tab)By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy (opens in a new tab), Terms of Service (opens in a new tab), and Dataset License (opens in a new tab)ACCEPT & CONTINUEor Only Accept Required

Just a moment...

a moment...Enable JavaScript and cookies to continue

Balance Bitter Flavor | Institute of Culinary Education

Skip to main site navigation

ApplyRequest InfoBlogCampusesRecreational ClassesHost Your EventUpcoming EventsClick to access the menu Close this panelSearch the Institute of Culinary Education Blog

GoCulinary ArtsPastry & Baking ArtsPlant-Based Culinary ArtsBusiness of FoodRecipesPeopleCultureVideos

Why Your Food Needs Bitterness Chef Jenny Dorsey makes the case against counteracting bitterness in food suggesting chefs master the nuance of layering the flavor.

Bitterness is a critical component of building intensity and complexity of flavor, but here in the U.S. we often shy away from using bitterness to accentuate our food and instead focus on tired adages like “fat is flavor” when attempting to build nuance into dishes. I’d like to divulge why bitterness is so key to developing complexity in finished foods, with a short guide on different ways you can play with the styles, levels and textures of the taste.

By Jenny Dorsey — ICE Chef

June 25, 2020Why is bitter an important part of flavor? To start, our taste buds are extremely attuned to the taste of bitterness, especially at the backs of our mouths. Scientists theorize this evolved as a “last chance” button of sorts to detect potentially poisonous foods and allow us to physically eject them from our bodies, as almost all toxic plants are bitter (which is also why poisonous substances like lead have been doubly harmful – not only do they attack our bodies but because they are sweet, children can voluntarily eat or lick lead-based paint). This means that we don’t need to use a lot of bitterness for most people to discern it on the plate – a little goes a long way and makes a big impact.Also very interestingly, bitter is one of our five primary tastes (six if you count fat, seven if you count spice), and it balances out umami – similar to how sourness helps neutralize fattiness. You can increase umami and bitterness as you cook, knowing that as long as the two are in sync, the final product will taste balanced. In my experience, incorporating a bitter note is a big part of the “craveability” factor that many of us chefs are striving for when we cook.Bitterness also helps cleanse our palates; in a way, it’s like taking a whiff of coffee beans in between wines during a tasting. When I think back to the times I’ve tasted something very delicious but just can’t stand more than a few bites before my taste buds are overwhelmed (e.g. most commonly when I eat truffles, wagyu or caviar), I find it’s because there’s just not enough bitterness present to make me want more. While it’s not the flavor you want to blanket the food, I liken it to little boosts of adrenaline for your taste buds to keep them stimulated while reformatting your palate back to neutral so the next food still feels interesting as we continue to eat. Maybe this is why naturally caffeinated items are bitter? If you’ve seen my last post about dividing spices into top, middle and base notes, you can apply that guiding theory to how you cook and layer flavors throughout the process. When cooking, think about when and how your bitter component is being incorporated, as that will likely determine when that same bitterness will spark your diner’s taste receptors. Sprinkling a warm, rich, bitter ingredient like cacao powder on top to finish a dish versus braising meat in a juicy but tannic red wine for hours feels different because the ingredient’s bitterness reaches our tongues in varying stages when eating. Also consider how that bitter ingredient is cut: large strips of collards as a side yields different results than having thin chiffonades incorporated throughout a dish.⁣When eating, consider what hits your tongue first vs. last. What flavors are present in each of these stages? Where the bitterness “sits” in a dish is a manifestation of how it independently smells and tastes (e.g. celery seed is a middle note, dill seed a base note) and when and how it was incorporated. Do you want subtle bitterness dispersed evenly throughout a soup or concentrated at the beginning or end of the eating experience? If you’re creating a plated dish, is there bitterness in every component or accents like a few dots of a fluid gel?I’ve put together a short chart of bitter ingredients that you can experiment with. It is certainly not exhaustive but will hopefully help turn your attention to some interesting new components to play with the next time you cook. Why not use a whole lemon when braising to infuse some of that bitter pith into the foundation of that dish? What about spiking a classic vodka sauce with a touch of bitter liqueur instead? Perhaps black tea can be the secret ingredient of your next spice blend! Once you master the nuance of bitterness, you’ll see the boundaries of deepening flavor profiles will expand exponentially.Gain the foundational cooking skills to layer flavor in Culinary Arts.Tags:Culinary EducationCookingICE ChefFood ScienceAdd new comment

You must have JavaScript enabled to use this form.

Your name

Subject

Comment

Leave this field blank

Related Stories02.28.24Culinary Arts8 Important American Cheesemakers You Should Know01.19.24Culinary ArtsUnderstanding PDO Designation in Cheese, and Why it Matters01.16.24Culinary ArtsHow & Why You Should Volunteer at ICEView All Posts

ICE is accredited by ACCSC and licensed by BPPE (in CA) and BPSS (in NY), and is not regulated in TX under Chapter 132 of the Texas Education Code.© 2023 Institute of Culinary Education. All Rights Reserved.

Terms of UsePrivacy PolicyDMCA PolicyJoin the ICE TeamApplicant Privacy StatementNY Career Catalog & BrochureLA Career Catalog & BrochureLA Annual Report and School Performance Fact Sheets Bureau for Private and Postsecondary Education (BPPE) Higher Education Emergency Relief FundNondiscrimination Statement & Title IX PolicyHigher Education Consumer Information Disclosures

Do Not Sell My Personal Information

im钱包App官网最新版下载 数字资产服务平台

比特派最新版本下载|bitter flavor

BITTER中文(简体)翻译：剑桥词典

What Is Bitterness?

Site Currently Unavailable

Page restricted | ScienceDirect

Bitter or not? BitterPredict, a tool for predicting taste from chemical structure | Scientific Reports

13 Healthiest Bitter Foods - Bitter Fruits and Vegetables for Digestive Health

Page restricted | ScienceDirect

[PDF] Traditional Chinese Bitter Flavor theory: Is there any relation with taste type II receptors? | Semantic Scholar

Just a moment...

Balance Bitter Flavor | Institute of Culinary Education

im钱包App官网最新版下载
数字资产服务平台