Reducing false positive rate of docking-based virtual screening by active learning

Abstract

Machine learning-based scoring functions (MLSFs) have become a very favorable alternative to classical scoring functions because of their potential superior screening performance. However, the information of negative data used to construct MLSFs was rarely reported in the literature, and meanwhile the putative inactive molecules recorded in existing databases usually have obvious bias from active molecules. Here we proposed an easy-to-use method named AMLSF that combines active learning using negative molecular selection strategies with MLSF, which can iteratively improve the quality of inactive sets and thus reduce the false positive rate of virtual screening. We chose energy auxiliary terms learning as the MLSF and validated our method on eight targets in the diverse subset of DUD-E. For each target, we screened the IterBioScreen database by AMLSF and compared the screening results with those of the four control models. The results illustrate that the number of active molecules in the top 1000 molecules identified by AMLSF was significantly higher than those identified by the control models. In addition, the free energy calculation results for the top 10 molecules screened out by the AMLSF, null model and control models based on DUD-E also proved that more active molecules can be identified, and the false positive rate can be reduced by AMLSF.

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

“…Accurate prediction of protein–ligand binding affinity remains one of the grand challenges of computational chemistry and biology. 1–3 With the ever increasing amount of high-resolution experimentally determined protein–ligand structures, 4 the binding affinity prediction methods have switched from physics-based 5–11 to empirical scoring functions 12–15 and knowledge-based, 16,17 and in the last decade to machine learning 18–27 and deep learning based methods. 28–38 Especially, deep learning is an end-to-end method that is ideally suited to find hidden nonlinear relationships 39 between 3D protein–ligand complex structures and binding affinity.…”

Correlation of protein binding pocket properties with hits’ chemistries used in generation of ultra-large virtual libraries

“…In two cases where RMSD exceeded 1Å (PDB:5i96 and 5vv0), all the differences were in the part of the molecules exposed to the solvent, while poses inside the pocket were determined with high accuracy. Although the structures used appeared to predict binding correctly when tested with "native" ligands, the presence of false positives in virtual screens is inevitable [33]. Testing binding properties experimentally for all identified hits was not possible in the context of this study.…”

Synthesis, In Vivo Anticonvulsant Activity Evaluation and In Silico Studies of Some Quinazolin-4(3H)-One Derivatives

“…To remove the false positive results, the use of more than one docking software and clustering of the predicted conformations are common practices in the field of in silico research of bioactive compounds. For the same reason, to improve the molecular docking studies, another additional method used in the modern drug design research is represented by the molecular dynamics studies [60][61][62][63][64][65].…”