Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI)
Abstract The identification of novel drug-target interactions (DTI) is critical to drug discovery and drug repurposing to address contemporary medical and public health challenges presented by emergent diseases. Historically, computational methods have framed DTI prediction as a binary classificatio...
Main Authors: | , , , , , , , , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2022-08-01
|
Series: | Scientific Reports |
Online Access: | https://doi.org/10.1038/s41598-022-16493-9 |
_version_ | 1811315432847572992 |
---|---|
author | Kevin Dick Daniel G. Kyrollos Eric D. Cosoreanu Joseph Dooley Joshua S. Fryer Shaun M. Gordon Nikhil Kharbanda Martin Klamrowski Patrick N. L. LaCasse Thomas F. Leung Muneeb A. Nasir Chang Qiu Aisha S. Robinson Derek Shao Boyan R. Siromahov Evening Starlight Christophe Tran Christopher Wang Yu-Kai Yang James R. Green |
author_facet | Kevin Dick Daniel G. Kyrollos Eric D. Cosoreanu Joseph Dooley Joshua S. Fryer Shaun M. Gordon Nikhil Kharbanda Martin Klamrowski Patrick N. L. LaCasse Thomas F. Leung Muneeb A. Nasir Chang Qiu Aisha S. Robinson Derek Shao Boyan R. Siromahov Evening Starlight Christophe Tran Christopher Wang Yu-Kai Yang James R. Green |
author_sort | Kevin Dick |
collection | DOAJ |
description | Abstract The identification of novel drug-target interactions (DTI) is critical to drug discovery and drug repurposing to address contemporary medical and public health challenges presented by emergent diseases. Historically, computational methods have framed DTI prediction as a binary classification problem (indicating whether or not a drug physically interacts with a given protein target); however, framing the problem instead as a regression-based prediction of the physiochemical binding affinity is more meaningful. With growing databases of experimentally derived drug-target interactions (e.g. Davis, Binding-DB, and Kiba), deep learning-based DTI predictors can be effectively leveraged to achieve state-of-the-art (SOTA) performance. In this work, we formulated a DTI competition as part of the coursework for a senior undergraduate machine learning course and challenged students to generate component DTI models that might surpass SOTA models and effectively combine these component models as part of a meta-model using the Reciprocal Perspective (RP) multi-view learning framework. Following 6 weeks of concerted effort, 28 student-produced component deep-learning DTI models were leveraged in this work to produce a new SOTA RP-DTI model, denoted the Meta Undergraduate Student DTI (MUSDTI) model. Through a series of experiments we demonstrate that (1) RP can considerably improve SOTA DTI prediction, (2) our new double-cold experimental design is more appropriate for emergent DTI challenges, (3) that our novel MUSDTI meta-model outperforms SOTA models, (4) that RP can improve upon individual models as an ensembling method, and finally, (5) RP can be utilized for low computation transfer learning. This work introduces a number of important revelations for the field of DTI prediction and sequence-based, pairwise prediction in general. |
first_indexed | 2024-04-13T11:30:05Z |
format | Article |
id | doaj.art-c56bec5467bc47b7832605e3bef1064c |
institution | Directory Open Access Journal |
issn | 2045-2322 |
language | English |
last_indexed | 2024-04-13T11:30:05Z |
publishDate | 2022-08-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Scientific Reports |
spelling | doaj.art-c56bec5467bc47b7832605e3bef1064c2022-12-22T02:48:35ZengNature PortfolioScientific Reports2045-23222022-08-0112111910.1038/s41598-022-16493-9Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI)Kevin Dick0Daniel G. Kyrollos1Eric D. Cosoreanu2Joseph Dooley3Joshua S. Fryer4Shaun M. Gordon5Nikhil Kharbanda6Martin Klamrowski7Patrick N. L. LaCasse8Thomas F. Leung9Muneeb A. Nasir10Chang Qiu11Aisha S. Robinson12Derek Shao13Boyan R. Siromahov14Evening Starlight15Christophe Tran16Christopher Wang17Yu-Kai Yang18James R. Green19Department of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityDepartment of Systems and Computer Engineering, Carleton UniversityAbstract The identification of novel drug-target interactions (DTI) is critical to drug discovery and drug repurposing to address contemporary medical and public health challenges presented by emergent diseases. Historically, computational methods have framed DTI prediction as a binary classification problem (indicating whether or not a drug physically interacts with a given protein target); however, framing the problem instead as a regression-based prediction of the physiochemical binding affinity is more meaningful. With growing databases of experimentally derived drug-target interactions (e.g. Davis, Binding-DB, and Kiba), deep learning-based DTI predictors can be effectively leveraged to achieve state-of-the-art (SOTA) performance. In this work, we formulated a DTI competition as part of the coursework for a senior undergraduate machine learning course and challenged students to generate component DTI models that might surpass SOTA models and effectively combine these component models as part of a meta-model using the Reciprocal Perspective (RP) multi-view learning framework. Following 6 weeks of concerted effort, 28 student-produced component deep-learning DTI models were leveraged in this work to produce a new SOTA RP-DTI model, denoted the Meta Undergraduate Student DTI (MUSDTI) model. Through a series of experiments we demonstrate that (1) RP can considerably improve SOTA DTI prediction, (2) our new double-cold experimental design is more appropriate for emergent DTI challenges, (3) that our novel MUSDTI meta-model outperforms SOTA models, (4) that RP can improve upon individual models as an ensembling method, and finally, (5) RP can be utilized for low computation transfer learning. This work introduces a number of important revelations for the field of DTI prediction and sequence-based, pairwise prediction in general.https://doi.org/10.1038/s41598-022-16493-9 |
spellingShingle | Kevin Dick Daniel G. Kyrollos Eric D. Cosoreanu Joseph Dooley Joshua S. Fryer Shaun M. Gordon Nikhil Kharbanda Martin Klamrowski Patrick N. L. LaCasse Thomas F. Leung Muneeb A. Nasir Chang Qiu Aisha S. Robinson Derek Shao Boyan R. Siromahov Evening Starlight Christophe Tran Christopher Wang Yu-Kai Yang James R. Green Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) Scientific Reports |
title | Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) |
title_full | Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) |
title_fullStr | Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) |
title_full_unstemmed | Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) |
title_short | Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI) |
title_sort | reciprocal perspective as a super learner improves drug target interaction prediction musdti |
url | https://doi.org/10.1038/s41598-022-16493-9 |
work_keys_str_mv | AT kevindick reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT danielgkyrollos reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT ericdcosoreanu reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT josephdooley reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT joshuasfryer reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT shaunmgordon reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT nikhilkharbanda reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT martinklamrowski reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT patricknllacasse reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT thomasfleung reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT muneebanasir reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT changqiu reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT aishasrobinson reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT derekshao reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT boyanrsiromahov reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT eveningstarlight reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT christophetran reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT christopherwang reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT yukaiyang reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti AT jamesrgreen reciprocalperspectiveasasuperlearnerimprovesdrugtargetinteractionpredictionmusdti |