Computational Drug Discovery Methods

The paper demonstrates the use of RetroPath2.0 for molecular structure enumeration and virtual screening in chemical space, including applications for glass transition temperature prediction and antibacterial activity optimization.

The paper introduces MOSAIC, a database and web resource for chemical-genetic interaction data, enabling mode-of-action prediction and discovery of compound effects in yeast.

The paper uses tensor decomposition on integrated gene expression data from heart failure and DrugMatrix datasets to identify drug candidates associated with heart failure.

PharmacoDB integrates large pharmacogenomic studies to enable comparison and analysis of drug screening data by curating identifiers and standardizing drug response metrics.

This paper introduces kernel multitask regression to predict compound toxicity on cell lines using chemical and genetic data, achieving strong performance on a benchmark challenge.

This paper uses a biomedical knowledge network to predict drug-disease treatment pairs and prioritize drug repurposing candidates.

This study identifies potential drug repositioning hypotheses by analyzing clinical phenotypes and prescription drug usage in a healthy population.

This paper introduces DTINet, a network integration method for predicting drug-target interactions and repositioning drugs, showing improved performance and validated novel COX inhibitor applications.

This study links polypharmacological drugs to CATH functional families, showing their network properties relate to drug side effects and target identification.

The paper integrates drug chemical structures with transcriptional responses to identify noisy responses and a transcriptional signature for phospholipidosis.

This study compares machine learning methods on a standardized ChEMBL dataset, finding deep neural networks outperform conventional methods in bioactivity prediction.

The paper introduces ProGENI, a gene prioritization method that improves drug sensitivity prediction by integrating gene expression and prior biological knowledge, identifying new genes involved in chemoresistance.

This paper introduces a 3D molecular representation (E3FP) that improves drug-to-target prediction compared to 2D fingerprints, with experimental validation of novel binding predictions.

This study examines tissue-specific drug sensitivity in in vitro datasets and finds strong associations between tissue origin and drug response, but most preclinical findings do not align with clinical observations.

This paper introduces DNF, a computational framework that classifies new drugs by integrating structural, perturbation, and sensitivity data to infer drug relationships and taxonomies.

This study uses interactome topology to predict drug side effects and shows that drug targets are closer to proteins causing known side effects.

The paper compares single-gene and multi-gene predictors of drug sensitivity in cancer cell lines using gene expression data and finds that multi-gene models perform better overall, though single-gene markers have higher precision.

This study compares pharmacophore-based and substructure-based alignment methods to identify structural requirements for antimalarial activity in substituted 1,2,4-trioxanes, finding pharmacophore-based alignment superior.