miDruglikeness: Subdivisional Drug-Likeness Prediction Models Using Active Ensemble Learning Strategies

Abstract

The drug development pipeline involves several stages including in vitro assays, in vivo assays, and clinical trials. For candidate selection, it is important to consider that a compound will successfully pass through these stages. Using graph neural networks, we developed three subdivisional models to individually predict the capacity of a compound to enter in vivo testing, clinical trials, and market approval stages. Furthermore, we proposed a strategy combing both active learning and ensemble learning to improve the quality of the models. The models achieved satisfactory performance in the internal test datasets and four self-collected external test datasets. We also employed the models as a general index to make an evaluation on a widely known benchmark dataset DEKOIS 2.0, and surprisingly found a powerful ability on virtual screening tasks. Our model system (termed as miDruglikeness) provides a comprehensive drug-likeness prediction tool for drug discovery and development.

Drug approval prediction based on the discrepancy in gene perturbation effects between cells and humans

“… 65 Using drug target information can help overcome the limitations of using chemical information in drug assessment. 64 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 In this study, we found that the performance of the drug approval prediction model was improved by using both types of information ( Fig. 5 b and c).…”

“… 63 Cai et al. 64 developed a machine-learning model based on chemical information to predict the ability of the drug to be approved. However, chemical information may have limitations in drug approval prediction because of the deviation of drug-likeness rules for newly developed drugs.…”

Current Trends and Challenges in Drug-Likeness Prediction: Are They Generalizable and Interpretable?

“…Cai et al. [ 29 ] developed a 3-subdivisional drug-likeness prediction model system, which consists of 3 individually trained models for evaluating the potential to reach in vivo, investigational, and approved stages progressively from in-stock compounds. They also combined active learning with ensemble learning to enhance the predictive ability of these models.…”

“…To address this challenge, some drug-likeness methods have employed subdivisional labels assigned by chemists [ 20 ] or incorporated research progress. [ 29 ] Some others only fit drugs to gain independence from the non-drug-like background, such as QED [ 53 ] and self-supervised RNN [ 30 ].…”

Advances in Drug Design and Development for Human Therapeutics Using Artificial Intelligence-II

“…Lipinski and coworkers introduced 'the Rule of 5' (Ro5) to exclude compounds with poor absorption or permeation, which are unfavorable for drug development. In this study, Cai et al [6] have introduced a graph neural network-based platform called miDruglikeness, which outperforms quantitative estimate of drug likeness (QED) by quickly assessing a vast amount of molecular data, including those generated by deep generative networks, offering improved predictive accuracy.…”