Back

A simple and efficient metric quantifying druggable property of chemical small molecules

Huang, C.; Yang, J.; Cui, Q.

2020-07-13 bioinformatics
10.1101/2020.07.13.199752 bioRxiv
Show abstract

One big class of drugs are chemical small molecules (CSMs), but the majority of CSMs are in very low druggable potential. Therefore, it is quite important to predict drug-related properties (druggable properties) for candidate CSMs. Currently, although a number of druggable properties (e.g. logP and pKa) can be calculated by in silico methods, the identification of druggable CSMs is still at high risk and new quantitative metrics for the druggable potential of CSMs are increasingly needed. Here, we presented normalized bond energy (NBE), a new metric for the above purpose. By applying NBE to the DrugBank CSMs whose properties are largely known, we revealed that NBE is able to describe a number of critical druggable properties including logP, pKa, membrane permeability, blood-brain barrier penetration, and human intestinal absorption. Moreover, given that the human endogenous metabolites could be served as an important resource for drug discovery, we applied NBE to the metabolites in the Human Metabolome Database. As a result, NBE shows a significant difference in metabolites from various body fluids and is correlated with some important properties including melting point and water solubility.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.