Towards a more inductive world for drug repurposing approaches

Drug–target interaction (DTI) prediction is a challenging albeit essential task in drug repurposing. Learning on graph models has drawn special attention as they can substantially reduce drug repurposing costs and time commitment. However, many current approaches require high-demand additional information besides DTIs that complicates their evaluation process and usability. Additionally, structural differences in the learning architecture of current models hinder their fair benchmarking. In this work, we first perform an in-depth evaluation of current DTI datasets and prediction models through a robust benchmarking process and show that DTI methods based on transductive models lack generalization and lead to inflated performance when traditionally evaluated, making them unsuitable for drug repurposing. We then propose a biologically driven strategy for negative-edge subsampling and uncovered previously unknown interactions via in vitro validation, missed by traditional subsampling. Finally, we provide a toolbox from all generated resources, crucial for fair benchmarking and robust model design. The authors address the challenge of predicting drug–target interactions, which is crucial for drug repurposing, by introducing a robust benchmarking framework. Using a biologically driven strategy, they uncover previously unknown interactions.

Authors

Institutions

Publication Details

Journal
Nature Machine Intelligence
Published
2025-02-13
DOI
https://doi.org/10.1038/s42256-025-00987-y
Citations
6
Primary Topic
Computational Drug Discovery Methods
Type
article
Field-Weighted Citation Impact
9.03

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Towards a more inductive world for drug repurposing approaches

Marija Pizurica, Mikel Hernáez, Idoia Ochoa, Silvestre Vicent et al.
6 citations
Nature Machine Intelligence
Computational Drug Discovery Methods
9.03
article

Towards a more inductive world for drug repurposing approaches

Marija Pizurica, Mikel Hernáez, Idoia Ochoa, Silvestre Vicent, Leonor Puchades‐Carrasco, Jesús de la Fuente, Laura Vera, Guillermo Serrano, Mikel Casals, Nuria Gómez-Cebrián, Antonio Pineda‐Lucena, Olivier Gevaert, Uxía Veleiro
article en
6 citations

Abstract

Drug–target interaction (DTI) prediction is a challenging albeit essential task in drug repurposing. Learning on graph models has drawn special attention as they can substantially reduce drug repurposing costs and time commitment. However, many current approaches require high-demand additional information besides DTIs that complicates their evaluation process and usability. Additionally, structural differences in the learning architecture of current models hinder their fair benchmarking. In this work, we first perform an in-depth evaluation of current DTI datasets and prediction models through a robust benchmarking process and show that DTI methods based on transductive models lack generalization and lead to inflated performance when traditionally evaluated, making them unsuitable for drug repurposing. We then propose a biologically driven strategy for negative-edge subsampling and uncovered previously unknown interactions via in vitro validation, missed by traditional subsampling. Finally, we provide a toolbox from all generated resources, crucial for fair benchmarking and robust model design. The authors address the challenge of predicting drug–target interactions, which is crucial for drug repurposing, by introducing a robust benchmarking framework. Using a biologically driven strategy, they uncover previously unknown interactions.

Nature Machine IntelligenceVol. 7(3)
Ghent University (BE), Navarre Institute of Health Research (ES), Navarrabiomed (ES), Gobierno de Navarra (ES), Clinica Universidad de Navarra (ES), Centro de Investigación Biomédica en Red de Cáncer (ES), Instituto de Investigación Sanitaria La Fe (ES), King Abdullah University of Science and Technology (SA), Universidad de Navarra (ES), Stanford University (US)
Fulbright Association, Ministerio de Ciencia, Innovación y Universidades, European Commission, Ministerio de Ciencia e Innovación, Consejería de Educación e Investigación, Gobierno de Navarra, Instituto de Salud Carlos III, Dirección General de Universidades e Investigación, Agencia Estatal de Investigación
Gender equality
Openalex Percentile: Top 2%
Computational Drug Discovery Methods
9.03
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.