Models and Processes to Extract Drug-like Molecules From Natural Language Text
Researchers worldwide are seeking to repurpose existing drugs or discover new drugs to counter the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A promising source of candidates for such studies is molecules that have been reported in the scientific literature to be...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2021-08-01
|
Series: | Frontiers in Molecular Biosciences |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fmolb.2021.636077/full |
id |
doaj-25dc6e150f3044b7a3c02a54d6b2bba8 |
---|---|
record_format |
Article |
spelling |
doaj-25dc6e150f3044b7a3c02a54d6b2bba82021-09-03T15:34:52ZengFrontiers Media S.A.Frontiers in Molecular Biosciences2296-889X2021-08-01810.3389/fmolb.2021.636077636077Models and Processes to Extract Drug-like Molecules From Natural Language TextZhi Hong0J. Gregory Pauloski1Logan Ward2Kyle Chard3Kyle Chard4Kyle Chard5Ben Blaiszik6Ben Blaiszik7Ian Foster8Ian Foster9Ian Foster10Department of Computer Science, University of Chicago, Chicago, IL, United StatesDepartment of Computer Science, University of Chicago, Chicago, IL, United StatesData Science and Learning Division, Argonne National Laboratory, Lemont, IL, United StatesDepartment of Computer Science, University of Chicago, Chicago, IL, United StatesData Science and Learning Division, Argonne National Laboratory, Lemont, IL, United StatesGlobus, University of Chicago, Chicago, IL, United StatesData Science and Learning Division, Argonne National Laboratory, Lemont, IL, United StatesGlobus, University of Chicago, Chicago, IL, United StatesDepartment of Computer Science, University of Chicago, Chicago, IL, United StatesData Science and Learning Division, Argonne National Laboratory, Lemont, IL, United StatesGlobus, University of Chicago, Chicago, IL, United StatesResearchers worldwide are seeking to repurpose existing drugs or discover new drugs to counter the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A promising source of candidates for such studies is molecules that have been reported in the scientific literature to be drug-like in the context of viral research. However, this literature is too large for human review and features unusual vocabularies for which existing named entity recognition (NER) models are ineffective. We report here on a project that leverages both human and artificial intelligence to detect references to such molecules in free text. We present 1) a iterative model-in-the-loop method that makes judicious use of scarce human expertise in generating training data for a NER model, and 2) the application and evaluation of this method to the problem of identifying drug-like molecules in the COVID-19 Open Research Dataset Challenge (CORD-19) corpus of 198,875 papers. We show that by repeatedly presenting human labelers only with samples for which an evolving NER model is uncertain, our human-machine hybrid pipeline requires only modest amounts of non-expert human labeling time (tens of hours to label 1778 samples) to generate an NER model with an F-1 score of 80.5%—on par with that of non-expert humans—and when applied to CORD’19, identifies 10,912 putative drug-like molecules. This enriched the computational screening team’s targets by 3,591 molecules, of which 18 ranked in the top 0.1% of all 6.6 million molecules screened for docking against the 3CLPro protein.https://www.frontiersin.org/articles/10.3389/fmolb.2021.636077/fullcoronavirus disease-19coronavirus disease-19 open research dataset challengenamed entity recognitionlong-short term memorydata mining |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Zhi Hong J. Gregory Pauloski Logan Ward Kyle Chard Kyle Chard Kyle Chard Ben Blaiszik Ben Blaiszik Ian Foster Ian Foster Ian Foster |
spellingShingle |
Zhi Hong J. Gregory Pauloski Logan Ward Kyle Chard Kyle Chard Kyle Chard Ben Blaiszik Ben Blaiszik Ian Foster Ian Foster Ian Foster Models and Processes to Extract Drug-like Molecules From Natural Language Text Frontiers in Molecular Biosciences coronavirus disease-19 coronavirus disease-19 open research dataset challenge named entity recognition long-short term memory data mining |
author_facet |
Zhi Hong J. Gregory Pauloski Logan Ward Kyle Chard Kyle Chard Kyle Chard Ben Blaiszik Ben Blaiszik Ian Foster Ian Foster Ian Foster |
author_sort |
Zhi Hong |
title |
Models and Processes to Extract Drug-like Molecules From Natural Language Text |
title_short |
Models and Processes to Extract Drug-like Molecules From Natural Language Text |
title_full |
Models and Processes to Extract Drug-like Molecules From Natural Language Text |
title_fullStr |
Models and Processes to Extract Drug-like Molecules From Natural Language Text |
title_full_unstemmed |
Models and Processes to Extract Drug-like Molecules From Natural Language Text |
title_sort |
models and processes to extract drug-like molecules from natural language text |
publisher |
Frontiers Media S.A. |
series |
Frontiers in Molecular Biosciences |
issn |
2296-889X |
publishDate |
2021-08-01 |
description |
Researchers worldwide are seeking to repurpose existing drugs or discover new drugs to counter the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A promising source of candidates for such studies is molecules that have been reported in the scientific literature to be drug-like in the context of viral research. However, this literature is too large for human review and features unusual vocabularies for which existing named entity recognition (NER) models are ineffective. We report here on a project that leverages both human and artificial intelligence to detect references to such molecules in free text. We present 1) a iterative model-in-the-loop method that makes judicious use of scarce human expertise in generating training data for a NER model, and 2) the application and evaluation of this method to the problem of identifying drug-like molecules in the COVID-19 Open Research Dataset Challenge (CORD-19) corpus of 198,875 papers. We show that by repeatedly presenting human labelers only with samples for which an evolving NER model is uncertain, our human-machine hybrid pipeline requires only modest amounts of non-expert human labeling time (tens of hours to label 1778 samples) to generate an NER model with an F-1 score of 80.5%—on par with that of non-expert humans—and when applied to CORD’19, identifies 10,912 putative drug-like molecules. This enriched the computational screening team’s targets by 3,591 molecules, of which 18 ranked in the top 0.1% of all 6.6 million molecules screened for docking against the 3CLPro protein. |
topic |
coronavirus disease-19 coronavirus disease-19 open research dataset challenge named entity recognition long-short term memory data mining |
url |
https://www.frontiersin.org/articles/10.3389/fmolb.2021.636077/full |
work_keys_str_mv |
AT zhihong modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT jgregorypauloski modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT loganward modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT kylechard modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT kylechard modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT kylechard modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT benblaiszik modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT benblaiszik modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT ianfoster modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT ianfoster modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext AT ianfoster modelsandprocessestoextractdruglikemoleculesfromnaturallanguagetext |
_version_ |
1717816130004844544 |