Computational Drug Discovery Methods

This paper introduces GraphDRP, a graph convolutional network for drug response prediction, showing improved performance and interpretability through molecular graph representation.

The paper introduces a computational method to identify drug repositioning candidates by overcoming biomedical network sparseness using compressed sensing.

The paper presents a machine learning framework to identify FDA-approved drugs for repurposing in Alzheimer's Disease by analyzing gene lists associated with disease severity and molecular mechanisms.

TranSynergy improves synergistic drug combination prediction by integrating biological knowledge and enhancing interpretability, revealing novel pathways and predicting new combinations for ovarian cancer.

This study uses tensor decomposition to analyze gene expression data and finds non-linear dose dependence can occur with single drugs, not just drug combinations.

PathDSP predicts drug sensitivity using cancer pathway enrichment, outperforming existing models and providing explainable results on GDSC and CCLE datasets.

AnnapuRNA is a scoring function for predicting RNA-small molecule interactions, outperforming existing tools by using statistical data from experimental RNA-ligand complexes.

This paper introduces DeepCE, a deep learning framework for chemical-induced gene expression prediction and compound screening, showing improved performance over existing methods and demonstrating its application to COVID-19 drug repurposing.

SSnet predicts protein-ligand interactions using secondary structure features, outperforming existing models and revealing key structural elements for ligand binding.

This paper introduces DeepCoy, a deep learning method to generate decoy molecules that match active molecules' properties, improving virtual screening performance by reducing bias and enhancing benchmarking.

The paper uses a Transformer neural network to generate novel drug molecules for a target protein based only on its amino acid sequence, producing drug-like compounds with plausible physicochemical properties.

The paper evaluates six web-based tools for predicting kinase mutation effects on ligand binding using a benchmark dataset, finding that some tools perform well when assessing directional changes in binding affinity.

The paper models SARS-CoV-2 Mpro substrate and ligand binding to identify new inhibitor opportunities through computational analysis of binding interactions.

MeFSAT is a curated database of secondary metabolites from medicinal fungi, providing chemical and therapeutic information to aid drug discovery.

DEELIG predicts protein-ligand binding affinity using deep learning without requiring docked poses or complexes as input.

This paper introduces DEVELOP, a generative model that incorporates 3D pharmacophoric constraints to improve molecule design quality and control.

This study identifies potential APOE4 modulators for coronary artery disease using molecular docking and simulation to evaluate ligand binding affinity and interactions.

This study evaluates the network pharmacology of Tephrosia purpurea phytochemicals against PCOS and its psychiatric comorbidities, identifying potential drug-like compounds and their molecular targets.

OCTAD enables virtual screening of therapeutics for cancer patient groups using gene expression features, prioritizing potent drug candidates for experimental testing.

This paper develops interpretable deep learning models for structure-free compound-protein affinity prediction and contact mapping, showing improved interpretability and performance.

The paper presents a Python program that screens over a billion compounds from PubChem using machine learning to generate drug leads for Dengue and West Nile viruses, followed by automated In Silico modelling of their interactions with methyltransferase targets.

This paper introduces dynamic hybrid pharmacophore models (DHPMs) to identify novel anti-TB chemotypes by screening compounds against multiple binding sites of Mtb-DapB, showing improved structural diversity and drug-like properties compared to conventional models.

Drug2ways predicts drug candidates by analyzing causal paths in biological networks for disease treatment and drug repositioning.

comboFM predicts drug combination effects using machine learning, achieving high accuracy in cancer cell line screening and validating novel synergies.

This paper improves molecular bioactivity prediction by introducing stochastic negative examples during training, increasing drug-screening benchmark performance by 121.7%.

This paper introduces a two-step multi-omics modeling approach to predict drug sensitivity in cancer cell lines, identifying key drivers of drug responsiveness and demonstrating improved performance over one-step methods.

The paper assesses structural similarities between common fluorophores and drugs, metabolites, and natural products, finding significant overlaps in chemical space.

This paper investigates the binding of phytocompounds from Rheum emodi to antibiotic resistance proteins and evaluates their drug-likeness and toxicity.

This paper introduces FL-QSAR, a federated learning framework for QSAR analysis, demonstrating its effectiveness in collaborative drug discovery while preserving privacy.