Abstract:Determining various properties of molecules is a critical step in drug discovery. Recently, with the improvement of large heterogeneous datasets and the development of deep learning approaches, more and more scientists have turned their attention to neural network-based virtual preliminary screening to reduce the time and monetary cost of drug discovery. However, the poor interpretability of deep learning masks causality, so models' conclusions are often beyond the comprehension of human users, which reduces the credibility of the model and makes it difficult for chemists to further narrow the huge chemical space based on models' results. Thus, this study develops a novel framework consisting of Graph Neural Networks for feature extraction, Curriculum-Based Learning Strategies for optimization, and a Learning Binary Neural Tree (LBNT) for prediction, to improve the performance of neural networks and reveal their decision-making process to chemists. The framework encodes molecular graph data with graph neural networks (GNNs), then retrains the encoder with curriculum-based learning strategies to reduce uncertainty and improve accuracy, and finally uses LBNT as the predictor, which joint retrains with the encoder after independently training, for prediction and visualization. The framework is validated on the public datasets and compared to single GNNs with normal training strategies as well as GNN encoders with common machine learning predictors instead of the LBNT predictor. The result reveals that the proposed framework enhances the point prediction accuracy of the completely trained GNN and reduces its uncertainty through curriculum-based learning, and further improves the accuracy by combining LBNT. Besides, compared with common machine learning tools, the LBNT predictor generally has the best performance because of joint retraining with the GNN encoder. The decision-making process of LBNT is also better and easier to explain than that of other models.

Conformalized Graph Learning for Molecular ADMET Property Prediction and Reliable Uncertainty Quantification.

Deep Learning-Based Conformal Prediction of Toxicity

An Adaptive Graph Learning Method for Automated Molecular Interactions and Properties Predictions

A Bayesian graph convolutional network for reliable prediction of molecular properties with uncertainty quantification

Uncertainty Quantification for Molecular Property Predictions with Graph Neural Architecture Search

Conformalized Link Prediction on Graph Neural Networks

HiGNN: Hierarchical Informative Graph Neural Networks for Molecular Property Prediction Equipped with Feature-Wise Attention

Prediction of Molecular Conformation using Deep Generative Neural Networks

Graph Neural Tree: A novel and interpretable deep learning-based framework for accurate molecular property predictions

ACR-GNN: Adaptive Cluster Reinforcement Graph Neural Network Based on Contrastive Learning

Chemi-net: a graph convolutional network for accurate drug property prediction

AEGNN-M:A 3D Graph-Spatial Co-Representation Model for Molecular Property Prediction

DropConn: Dropout Connection Based Random GNNs for Molecular Property Prediction

Enhanced Graph Isomorphism Network For Molecular Admet Properties Prediction

Explaining Graph Neural Networks with Large Language Models: A Counterfactual Perspective for Molecular Property Prediction

Evaluating Scalable Uncertainty Estimation Methods for Deep Learning-Based Molecular Property Prediction

Advanced graph and sequence neural networks for molecular property prediction and drug discovery

Uncertainty Quantification Using Neural Networks for Molecular Property Prediction

Ligand binding affinity prediction with fusion of graph neural networks and 3D structure-based complex graph

Improved Uncertainty Estimation of Graph Neural Network Potentials Using Engineered Latent Space Distances

Development and Evaluation of Conformal Prediction Methods for QSAR