3D Deep Learning for Biological Function Prediction from Physical Fields

Predicting the biological function of molecules, be it proteins or drug-like compounds, from their atomic structure is an important and long-standing problem. Function is dictated by structure, since it is by spatial interactions that molecules interact with each other, both in terms of steric compl...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Golkov, Vladimir, Skwark, Marcin J, Mirchev, Atanas, Dikov, Georgi, Geanes, Alexander R, Mendenhall, Jeffrey, Meiler, Jens, Cremers, Daniel
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Golkov, Vladimir
Skwark, Marcin J
Mirchev, Atanas
Dikov, Georgi
Geanes, Alexander R
Mendenhall, Jeffrey
Meiler, Jens
Cremers, Daniel
description Predicting the biological function of molecules, be it proteins or drug-like compounds, from their atomic structure is an important and long-standing problem. Function is dictated by structure, since it is by spatial interactions that molecules interact with each other, both in terms of steric complementarity, as well as intermolecular forces. Thus, the electron density field and electrostatic potential field of a molecule contain the "raw fingerprint" of how this molecule can fit to binding partners. In this paper, we show that deep learning can predict biological function of molecules directly from their raw 3D approximated electron density and electrostatic potential fields. Protein function based on EC numbers is predicted from the approximated electron density field. In another experiment, the activity of small molecules is predicted with quality comparable to state-of-the-art descriptor-based methods. We propose several alternative computational models for the GPU with different memory and runtime requirements for different sizes of molecules and of databases. We also propose application-specific multi-channel data representations. With future improvements of training datasets and neural network settings in combination with complementary information sources (sequence, genomic context, expression level), deep learning can be expected to show its generalization power and revolutionize the field of molecular function prediction.
doi_str_mv 10.48550/arxiv.1704.04039
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1704_04039</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1704_04039</sourcerecordid><originalsourceid>FETCH-LOGICAL-a679-78b5282dfc3d5d04a3fb9b7a4260a07e5a5d221c2b45d4aedfa495a96759cd203</originalsourceid><addsrcrecordid>eNotz71OwzAUBWAvDKjlAZjwCyTc-CeOx9JSihSpHbpHN752sZTGlQOIvj2iYTpnODrSx9hjBaVqtIZnzD_xu6wMqBIUSHvPdnLDN95feOsxj3E88ZAyf4lpSKfocODbr9F9xjTyQ_YU5xpyOvPDx3WaF9EPNC3ZXcBh8g__uWDH7etxvSva_dv7etUWWBtbmKbXohEUnCRNoFCG3vYGlagBwXiNmoSonOiVJoWeAiqr0dZGW0cC5II9zbc3SnfJ8Yz52v2RuhtJ_gIh0UZC</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>3D Deep Learning for Biological Function Prediction from Physical Fields</title><source>arXiv.org</source><creator>Golkov, Vladimir ; Skwark, Marcin J ; Mirchev, Atanas ; Dikov, Georgi ; Geanes, Alexander R ; Mendenhall, Jeffrey ; Meiler, Jens ; Cremers, Daniel</creator><creatorcontrib>Golkov, Vladimir ; Skwark, Marcin J ; Mirchev, Atanas ; Dikov, Georgi ; Geanes, Alexander R ; Mendenhall, Jeffrey ; Meiler, Jens ; Cremers, Daniel</creatorcontrib><description>Predicting the biological function of molecules, be it proteins or drug-like compounds, from their atomic structure is an important and long-standing problem. Function is dictated by structure, since it is by spatial interactions that molecules interact with each other, both in terms of steric complementarity, as well as intermolecular forces. Thus, the electron density field and electrostatic potential field of a molecule contain the "raw fingerprint" of how this molecule can fit to binding partners. In this paper, we show that deep learning can predict biological function of molecules directly from their raw 3D approximated electron density and electrostatic potential fields. Protein function based on EC numbers is predicted from the approximated electron density field. In another experiment, the activity of small molecules is predicted with quality comparable to state-of-the-art descriptor-based methods. We propose several alternative computational models for the GPU with different memory and runtime requirements for different sizes of molecules and of databases. We also propose application-specific multi-channel data representations. With future improvements of training datasets and neural network settings in combination with complementary information sources (sequence, genomic context, expression level), deep learning can be expected to show its generalization power and revolutionize the field of molecular function prediction.</description><identifier>DOI: 10.48550/arxiv.1704.04039</identifier><language>eng</language><subject>Computer Science - Learning ; Quantitative Biology - Biomolecules ; Quantitative Biology - Quantitative Methods ; Statistics - Machine Learning</subject><creationdate>2017-04</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,777,882</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1704.04039$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1704.04039$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Golkov, Vladimir</creatorcontrib><creatorcontrib>Skwark, Marcin J</creatorcontrib><creatorcontrib>Mirchev, Atanas</creatorcontrib><creatorcontrib>Dikov, Georgi</creatorcontrib><creatorcontrib>Geanes, Alexander R</creatorcontrib><creatorcontrib>Mendenhall, Jeffrey</creatorcontrib><creatorcontrib>Meiler, Jens</creatorcontrib><creatorcontrib>Cremers, Daniel</creatorcontrib><title>3D Deep Learning for Biological Function Prediction from Physical Fields</title><description>Predicting the biological function of molecules, be it proteins or drug-like compounds, from their atomic structure is an important and long-standing problem. Function is dictated by structure, since it is by spatial interactions that molecules interact with each other, both in terms of steric complementarity, as well as intermolecular forces. Thus, the electron density field and electrostatic potential field of a molecule contain the "raw fingerprint" of how this molecule can fit to binding partners. In this paper, we show that deep learning can predict biological function of molecules directly from their raw 3D approximated electron density and electrostatic potential fields. Protein function based on EC numbers is predicted from the approximated electron density field. In another experiment, the activity of small molecules is predicted with quality comparable to state-of-the-art descriptor-based methods. We propose several alternative computational models for the GPU with different memory and runtime requirements for different sizes of molecules and of databases. We also propose application-specific multi-channel data representations. With future improvements of training datasets and neural network settings in combination with complementary information sources (sequence, genomic context, expression level), deep learning can be expected to show its generalization power and revolutionize the field of molecular function prediction.</description><subject>Computer Science - Learning</subject><subject>Quantitative Biology - Biomolecules</subject><subject>Quantitative Biology - Quantitative Methods</subject><subject>Statistics - Machine Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz71OwzAUBWAvDKjlAZjwCyTc-CeOx9JSihSpHbpHN752sZTGlQOIvj2iYTpnODrSx9hjBaVqtIZnzD_xu6wMqBIUSHvPdnLDN95feOsxj3E88ZAyf4lpSKfocODbr9F9xjTyQ_YU5xpyOvPDx3WaF9EPNC3ZXcBh8g__uWDH7etxvSva_dv7etUWWBtbmKbXohEUnCRNoFCG3vYGlagBwXiNmoSonOiVJoWeAiqr0dZGW0cC5II9zbc3SnfJ8Yz52v2RuhtJ_gIh0UZC</recordid><startdate>20170413</startdate><enddate>20170413</enddate><creator>Golkov, Vladimir</creator><creator>Skwark, Marcin J</creator><creator>Mirchev, Atanas</creator><creator>Dikov, Georgi</creator><creator>Geanes, Alexander R</creator><creator>Mendenhall, Jeffrey</creator><creator>Meiler, Jens</creator><creator>Cremers, Daniel</creator><scope>AKY</scope><scope>ALC</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20170413</creationdate><title>3D Deep Learning for Biological Function Prediction from Physical Fields</title><author>Golkov, Vladimir ; Skwark, Marcin J ; Mirchev, Atanas ; Dikov, Georgi ; Geanes, Alexander R ; Mendenhall, Jeffrey ; Meiler, Jens ; Cremers, Daniel</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a679-78b5282dfc3d5d04a3fb9b7a4260a07e5a5d221c2b45d4aedfa495a96759cd203</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Computer Science - Learning</topic><topic>Quantitative Biology - Biomolecules</topic><topic>Quantitative Biology - Quantitative Methods</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Golkov, Vladimir</creatorcontrib><creatorcontrib>Skwark, Marcin J</creatorcontrib><creatorcontrib>Mirchev, Atanas</creatorcontrib><creatorcontrib>Dikov, Georgi</creatorcontrib><creatorcontrib>Geanes, Alexander R</creatorcontrib><creatorcontrib>Mendenhall, Jeffrey</creatorcontrib><creatorcontrib>Meiler, Jens</creatorcontrib><creatorcontrib>Cremers, Daniel</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv Quantitative Biology</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Golkov, Vladimir</au><au>Skwark, Marcin J</au><au>Mirchev, Atanas</au><au>Dikov, Georgi</au><au>Geanes, Alexander R</au><au>Mendenhall, Jeffrey</au><au>Meiler, Jens</au><au>Cremers, Daniel</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>3D Deep Learning for Biological Function Prediction from Physical Fields</atitle><date>2017-04-13</date><risdate>2017</risdate><abstract>Predicting the biological function of molecules, be it proteins or drug-like compounds, from their atomic structure is an important and long-standing problem. Function is dictated by structure, since it is by spatial interactions that molecules interact with each other, both in terms of steric complementarity, as well as intermolecular forces. Thus, the electron density field and electrostatic potential field of a molecule contain the "raw fingerprint" of how this molecule can fit to binding partners. In this paper, we show that deep learning can predict biological function of molecules directly from their raw 3D approximated electron density and electrostatic potential fields. Protein function based on EC numbers is predicted from the approximated electron density field. In another experiment, the activity of small molecules is predicted with quality comparable to state-of-the-art descriptor-based methods. We propose several alternative computational models for the GPU with different memory and runtime requirements for different sizes of molecules and of databases. We also propose application-specific multi-channel data representations. With future improvements of training datasets and neural network settings in combination with complementary information sources (sequence, genomic context, expression level), deep learning can be expected to show its generalization power and revolutionize the field of molecular function prediction.</abstract><doi>10.48550/arxiv.1704.04039</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.1704.04039
ispartof
issn
language eng
recordid cdi_arxiv_primary_1704_04039
source arXiv.org
subjects Computer Science - Learning
Quantitative Biology - Biomolecules
Quantitative Biology - Quantitative Methods
Statistics - Machine Learning
title 3D Deep Learning for Biological Function Prediction from Physical Fields
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-20T00%3A51%3A35IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=3D%20Deep%20Learning%20for%20Biological%20Function%20Prediction%20from%20Physical%20Fields&rft.au=Golkov,%20Vladimir&rft.date=2017-04-13&rft_id=info:doi/10.48550/arxiv.1704.04039&rft_dat=%3Carxiv_GOX%3E1704_04039%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true