Finding new anticancer medications has been generally concerned and stays an open test. Both phenotypic-based experimental screening and target-based experimental screening are common approaches to the discovery of anticancer drugs. Both of these approaches are time-consuming, labor-intensive, and expensive. In this review, we gathered 485,900 mixtures including in 3,919,974 bioactivity records against 426 anticancer targets and 346 disease cell lines from scholastic writing, as well as 60 growth cell lines from NCI-60 board. The FP-GNN deep learning method was then used to create a total of 832 classification models, including 426 target- and 406 cell-based predictive models, to predict the inhibitory activity of compounds against targets and tumor cell lines. The FP-GNN models outperform conventional machine learning and deep learning in terms of overall predictive performance, achieving the highest AUC values of 0.91, 0.88, and 0.91 for the test sets of targets, academia-sourced cancer cell lines, and NCI-60 cancer cell lines, respectively. Based on these high-quality models, the user-friendly webserver Deep Cancer Map and its local version made it possible for users to perform anticancer drug discovery tasks like large-scale virtual screening, profiling prediction of anticancer agents, target fishing, and drug repositioning. We guess this stage to speed up the disclosure of anticancer medications in the field.
HTML PDFShare this article
Research and Reports in Medical Sciences received 13 citations as per Google Scholar report
