Hierarchical multi-omics data integration and modeling predict cell-specific chemical proteomics and drug responses

Drug-induced phenotypes result from biomolecular interactions across various levels of a biological system. Characterization of pharmacological actions therefore requires integration of multi-omics data. Proteomics profiles, which may more directly reflect disease mechanisms and biomarkers than transcriptomics, have not been widely exploited due to data scarcity and frequent missing values. A computational method for inferring drug-induced proteome patterns would therefore enable progress in systems pharmacology. To predict the proteome profiles and corresponding phenotypes of an uncharacterized cell or tissue type that has been disturbed by an uncharacterized chemical, we developed an end-to-end deep learning framework: TransPro. TransPro hierarchically integrated multi-omics data, in line with the central dogma of molecular biology. Our in-depth assessments of TransPro's predictions of anti-cancer drug sensitivity and drug adverse reactions reveal that TransPro's accuracy is on par with that of experimental data. Hence, TransPro may facilitate the imputation of proteomics data and compound screening in systems pharmacology.

Mutation-oriented profiling of autoinhibitory kinase conformations predicts RAF inhibitor efficacies

Kinase-targeted therapies have the potential to improve the survival of patients with cancer. However, the cancer-specific spectrum of kinase alterations exhibits distinct functional properties and requires mutation-oriented drug treatments. Besides post-translational modifications and diverse intermolecular interactions of kinases, it is the distinct disease mutation which reshapes full-length kinase conformations, affecting their activity. Oncokinase mutation profiles differ between cancer types, as it was shown for BRAF in melanoma and non-small-cell lung cancers. Here, we present the target-oriented application of a kinase conformation (KinCon) reporter platform for live-cell measurements of autoinhibitory kinase activity states. The bioluminescence-based KinCon biosensor allows the tracking of conformation dynamics of full-length kinases in intact cells and real time. We show that the most frequent BRAF cancer mutations affect kinase conformations and thus the engagement and efficacy of V600E-specific BRAF inhibitors (BRAFi). We illustrate that the patient mutation harboring KinCon reporters display differences in the effectiveness of the three clinically approved BRAFi vemurafenib, encorafenib, and dabrafenib and the preclinical paradox breaker PLX8394. We confirmed KinCon-based drug efficacy predictions for BRAF mutations other than V600E in proliferation assays using patient-derived lung cancer cell lines and by analyzing downstream kinase signaling. The systematic implementation of such conformation reporters will allow to accelerate the decision process for the mutation-oriented RAF-kinase cancer therapy. Moreover, we illustrate that the presented kinase reporter concept can be extended to other kinases which harbor patient mutations. Overall, KinCon profiling provides additional mechanistic insights into full-length kinase functions by reporting protein-protein interaction (PPI)-dependent, mutation-specific, and drug-driven changes of kinase activity conformations.

MEDICASCY: A Machine Learning Approach for Predicting Small-Molecule Drug Side Effects, Indications, Efficacy, and Modes of Action

To improve the drug discovery yield, a method which is implemented at the beginning of drug discovery that accurately predicts drug side effects, indications, efficacy, and mode of action based solely on the input of the drug's chemical structure is needed. In contrast, extant predictive methods do not comprehensively address these aspects of drug discovery and rely on features derived from extensive, often unavailable experimental information for novel molecules. To address these issues, we developed MEDICASCY, a multilabel-based boosted random forest machine learning method that only requires the small molecule's chemical structure for the drug side effect, indication, efficacy, and probable mode of action target predictions; however, it has comparable or even significantly better performance than existing approaches requiring far more information. In retrospective benchmarking on high confidence predictions, MEDICASCY shows about 78% precision and recall for predicting at least one severe side effect and 72% precision drug efficacy. Experimental validation of MEDICASCY's efficacy predictions on novel molecules shows close to 80% precision for the inhibition of growth in ovarian, breast, and prostate cancer cell lines. Thus, MEDICASCY should improve the success rate for new drug approval. A web service for academic users is available at http://pwp.gatech.edu/cssb/MEDICASCY.