DebiasedDTA: A Framework for Improving the Generalizability of Drug-Target Affinity Prediction Models (2021-07-04T00:00:00.000000Z)

TL;DR

The results suggest that models trained in the DebiasedDTA framework can achieve improved generalizability in predicting the interactions of the previously unseen biomolecules, as well as performance improvements on those previously seen.

Abstract

Computational models that accurately predict the binding affinity of an input protein-chemical pair can accelerate drug discovery studies. These models are trained on available protein-chemical interaction datasets, which may contain dataset biases that may lead the model to learn dataset-specific patterns, instead of generalizable relationships. As a result, the prediction performance of models drops for previously unseen biomolecules, $\textit{i.e.}$ the prediction models cannot generalize to biomolecules outside of the dataset. The latest approaches that aim to improve model generalizability either have limited applicability or introduce the risk of degrading prediction performance. Here, we present DebiasedDTA, a novel drug-target affinity (DTA) prediction model training framework that addresses dataset biases to improve the generalizability of affinity prediction models. DebiasedDTA reweights the training samples to mitigate the effect of dataset biases and is applicable to most DTA prediction models. The results suggest that models trained in the DebiasedDTA framework can achieve improved generalizability in predicting the interactions of the previously unseen biomolecules, as well as performance improvements on those previously seen. Extensive experiments with different biomolecule representations, model architectures, and datasets demonstrate that DebiasedDTA can upgrade DTA prediction models irrespective of the biomolecule representation, model architecture, and training dataset. Last but not least, we release DebiasedDTA as an open-source python library to enable other researchers to debias their own predictors and/or develop their own debiasing methods. We believe that this python library will corroborate and foster research to develop more generalizable DTA prediction models.

Authors

References39 items

DrugOOD: Out-of-Distribution (OOD) Dataset Curator and Benchmark for AI-aided Drug Discovery - A Focus on Affinity Prediction Problems with Noise Annotations

MGraphDTA: deep multiscale graph neural network for explainable drug–target binding affinity prediction

Towards Out-Of-Distribution Generalization: A Survey

WILDS: A Benchmark of in-the-Wild Distribution Shifts

ChemBERTa: Large-Scale Self-Supervised Pretraining for Molecular Property Prediction

Using Single Protein/Ligand Binding Models to Predict Active Ligands for Unseen Proteins

ProtTrans: Towards Cracking the Language of Life’s Code Through Self-Supervised Deep Learning and High Performance Computing

In Search of Lost Domain Generalization

Noise or Signal: The Role of Image Backgrounds in Object Recognition

The Pitfalls of Simplicity Bias in Neural Networks

Shortcut learning in deep neural networks

LIT-PCBA: An Unbiased Data Set for Machine Learning and Virtual Screening

Dataset Augmentation Allows Deep Learning-Based Virtual Screening To Better Generalize To Unseen Target Classes, And Highlight Important Binding Interactions

Predicting or Pretending: Artificial Intelligence for Protein-Ligand Interactions Lack of Sufficiently Large and Unbiased Datasets

Correlation-aware Adversarial Domain Adaptation and Generalization

Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization

Invariant Risk Minimization

Learning from the ligand: using ligand-based features to improve binding affinity prediction

Hidden bias in the DUD-E dataset leads to misleading performance of deep learning in structure-based virtual screening

In Need of Bias Control: Evaluating Chemical Data for Machine Learning in Structure-Based Virtual Screening

A Review of Domain Adaptation without Target Labels

ChemBoost: A Chemical Language Based Approach for Protein – Ligand Binding Affinity Prediction

Hypothesis Only Baselines in Natural Language Inference

Annotation Artifacts in Natural Language Inference Data

DeepDTA: deep drug–target binding affinity prediction

Elements of Causal Inference: Foundations and Learning Algorithms

Benchmark of four popular virtual screening programs: construction of the active/decoy dataset remains a major determinant of measured performance

Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization

Neural Machine Translation of Rare Words with Subword Units

Discriminative Chemical Patterns: Automatic and Interactive Design

Making Sense of Large-Scale Kinase Inhibitor Bioactivity Data Sets: A Comparative and Integrative Analysis

Data completeness—the Achilles heel of drug-target networks

BindingDB: a web-accessible database of experimentally determined protein–ligand binding affinities

Concordance probability and discriminatory power in proportional hazards regression

The Effect of Debiasing Protein-Ligand Binding Data on Generalization

Most ligand-based classiﬁcation benchmarks reward memorization rather than generalization

Reproducible Drug Repurposing: When Similarity Does Not Suffice

Toward more realistic drug^target interaction predictions

GraphDTA: Predicting drug–target binding aﬃnity with graph neural networks