TranScreen: Transfer Learning on Graph-Based Anti-Cancer Virtual Screening Model

Deep learning’s automatic feature extraction has proven its superior performance over traditional fingerprint-based features in the implementation of virtual screening models. However, these models face multiple challenges in the field of early drug discovery, such as over-training and generalizatio...

Full description

Saved in:
Bibliographic Details
Published in:Big data and cognitive computing 2020-09, Vol.4 (3), p.16
Main Authors: Salem, Milad, Khormali, Aminollah, Arshadi, Arash Keshavarzi, Webb, Julia, Yuan, Jiann-Shiun
Format: Article
Language:English
Subjects:
Applied research
Artificial neural networks
Cancer
Cancer screening
Data processing
Datasets
Deep learning
Drug discovery
Feature extraction
Kinases
Machine learning
Methods
Neural networks
Screening
Technology application
transfer learning
virtual screening
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Deep learning’s automatic feature extraction has proven its superior performance over traditional fingerprint-based features in the implementation of virtual screening models. However, these models face multiple challenges in the field of early drug discovery, such as over-training and generalization to unseen data, due to the inherently unbalanced and small datasets. In this work, the TranScreen pipeline is proposed, which utilizes transfer learning and a collection of weight initializations to overcome these challenges. An amount of 182 graph convolutional neural networks are trained on molecular source datasets and the learned knowledge is transferred to the target task for fine-tuning. The target task of p53-based bioactivity prediction, an important factor for anti-cancer discovery, is chosen to showcase the capability of the pipeline. Having trained a collection of source models, three different approaches are implemented to compare and rank them for a given task before fine-tuning. The results show improvement in performance of the model in multiple cases, with the best model increasing the area under receiver operating curve ROC-AUC from 0.75 to 0.91 and the recall from 0.25 to 1. This improvement is vital for practical virtual screening via lowering the false negatives and demonstrates the potential of transfer learning. The code and pre-trained models are made accessible online.
ISSN:2504-2289
2504-2289
DOI:10.3390/bdcc4030016