Computational Drug Discovery Methods

Deep learning based predictive modeling to screen natural compounds against TNF-alpha for the potential management of Rheumatoid Arthritis: Virtual screening to comprehensive in silico investigation

This paper uses deep learning to screen natural compounds against TNF-alpha for RA treatment, identifying three potential inhibitors through in silico analysis.

DeepMol: An Automated Machine and Deep Learning Framework for Computational Chemistry

DeepMol automates the ML pipeline for molecular property prediction, achieving competitive results on benchmark datasets without manual feature engineering.

DRUGSYNC: Prediction of Synergistic Drug Combinations Using GCN Graph Convolutional Network based and Pre-learned Drug-induced Gene Expression Profiles

DRUGSYNC predicts synergistic drug combinations using pre-learned gene expression profiles and a GCN model, achieving high accuracy for unseen drug pairs.

Drug Response Modeling across Cancers: Proteomics vs. Transcriptomics

This study compares proteomics and transcriptomics for drug response prediction in cancer cell lines, finding proteomics provides better performance.

Biological Database Mining for LLM-Driven Alzheimer's Disease Drug Repurposing

This paper develops an LLM-driven pipeline for Alzheimer's drug repurposing by integrating ontologies and drug databases, showing improved alignment with expert knowledge and reduced hallucinations.

MDMD: a computational model for predicting drug-related microbes based on the aggregated metapaths from a heterogeneous network

The paper introduces MDMD, a model that predicts drug-related microbes using metapaths from a heterogeneous network, achieving 1.9% higher accuracy than existing methods.

Interpretable Drug Response Prediction through Molecule Structure-aware and Knowledge-Guided Visible Neural Network

DrugVNN predicts drug responses with high accuracy and interpretability by integrating gene features and drug representations using a novel attention mechanism.

Discovery of natural compounds as novel FMS-like tyrosine kinase-3 (FLT3) therapeutic inhibitors for the treatment of acute myeloid leukemia: an in silico approach

The paper identifies three natural compounds as potential FLT3 inhibitors for AML using computational methods including pharmacophore screening, docking, ADMET analysis, and molecular dynamics.

Structural Biophysics-Guided Computational Design of Semaglutide Analogues to Enhance GLP-1R Activation

This paper computationally designs semaglutide analogues to enhance GLP-1R activation using structural biophysics, resulting in 564 analogues with improved binding affinity.

Knowledge Graph and Machine Learning Help the Research of Drugs Aimed at Neurological Diseases

This study introduces PATHOS, a knowledge graph, and LOGOS, a graph embedding model, to advance drug research for neurological diseases through three case studies.

Receptor cavity-based screening reveals potential allosteric modulators of gonadotropin receptors in carp (Cyprinus carpio).

This study identifies allosteric modulators of carp gonadotropin receptors using computational screening and in vitro validation.

SurfDock is a Surface-Informed Diffusion Generative Model for Reliable and Accurate Protein-ligand Complex Prediction

SurfDock improves protein-ligand complex prediction by integrating multiple protein representations into a diffusion model, achieving higher accuracy and physical plausibility than existing methods.

Optimization of a Multi-Feature AI Ensemble and Voting System for MAPKAPK2 Inhibitor Discovery

The paper develops an ensemble deep learning model with voting systems to identify MAPKAPK2 inhibitors, achieving high accuracy and low false positive rates.

Ultra-high-throughput screening of antimicrobial combination therapies using a two-stage transparent machine learning model

This paper introduces M2D2, a two-stage ML model that identifies synergistic antimicrobial drug combinations and validates them experimentally against E. coli.

Auto QSAR-based Active learning docking for hit identification of potential inhibitors ofPlasmodium falciparum Hsp90 as antimalarial agents

The paper uses active learning and docking to identify potential inhibitors of Plasmodium falciparum Hsp90, finding FTN-T5 as a potent compound with moderate selectivity.

Deep Learning Unravels Potential Antibiotic-Resistance Drugs against Klebsiella Pneumoniae

This study uses deep learning to identify potential antibiotic-resistant drug candidates against Klebsiella pneumoniae with 72% prediction accuracy.

iSIM-sigma: efficient standard deviation calculation for molecular similarity

The paper introduces efficient methods to calculate standard deviation of molecular similarities with reduced computational complexity.

MetaGIN: A Lightweight Framework for Molecular Property Prediction

MetaGIN predicts molecular properties using a lightweight framework that incorporates 2-hop and 3-hop edges, achieving high accuracy with fewer parameters.

ApoDock: Ligand-Conditioned Sidechain Packing for Flexible Molecular Docking

ApoDock improves molecular docking by combining machine learning-driven sidechain packing with traditional sampling to generate physically plausible protein-ligand poses.

Representation Learning of Human Diseases for Indication Expansion and Investment Decisions

This paper introduces Dis2Vec, a representation learning framework for human diseases that enables therapeutic transferability prediction and disease clustering, enhancing translational research efficiency.

ProAffinity-GNN: A Novel Approach to Structure-based Protein-Protein Binding Affinity Prediction via a Curated Dataset and Graph Neural Networks

This paper introduces ProAffinity-GNN, a deep learning framework for predicting protein-protein binding affinities using graph neural networks and a curated dataset of protein complexes with experimentally determined affinities.

Deep Kernel Inversion: Rapid and Accurate Molecular Interaction Prediction for Drug Design

This paper introduces a deep learning framework for predicting molecular interactions, reducing computational complexity and demonstrating high performance in human protein-protein interaction mapping.

vScreenML v2.0: Improved Machine Learning Classification for Reducing False Positives in Structure-Based Virtual Screening

vScreenML v2.0 improves virtual screening by reducing false positives through a streamlined Python implementation, outperforming existing tools in hit discovery.

High-Throughput Screening Reveals Potential Inhibitors Targeting Trimethoprim-Resistant DfrA1 Protein in Klebsiella pneumoniae and Escherichia coli

This study identifies six potential inhibitors of the DfrA1 protein through high-throughput screening of 3,601 compounds, with DC4 and DC6 showing superior efficacy compared to the control drug.

Deep phenotypic profiling of neuroactive drugs in larval zebrafish

The paper trains a deep metric learning model on zebrafish behavioral profiles to identify neuroactive compounds, achieving better performance than correlation distance and validating predictions with in vitro assays.

Efficient Generation of Protein Pockets with PocketGen

PocketGen generates high-fidelity protein pockets with superior binding affinity and structural validity, operating ten times faster than physics-based methods.

Analysis of Curcumin and Piperine for Breast Cancer Prevention and Treatment via Simulation and Artificial Intelligence Methods

This paper uses simulation and AI to analyze Curcumin and Piperine for breast cancer prevention, finding they are synergistic and interact with TRPV1 and STAT3.

Pan-Cancer Drug Sensitivity Prediction from Gene Expression using Deep Learning

This paper develops a deep learning model to predict drug sensitivity in cancer using gene expression data and perturbation signatures, validated in prostate cancer cell lines.

Host-Guest binding free energies a la carte: an automated OneOPES protocol

This paper introduces an automated OneOPES protocol to improve binding free energy estimation in host-guest systems by enhancing sampling and enabling systematic force field comparison.

BEGAN: Boltzmann-Reweighted Data Augmentation for Enhanced GAN-Based Molecule Design in Insect Pheromone Receptors

This paper introduces BEGAN, a GAN-based method that uses Boltzmann reweighting to enhance molecule generation for improved binding to insect pheromone receptors.