drugdesign.org source of truth
-
Updated
Feb 18, 2024
drugdesign.org source of truth
This repository contains all the projects related to Chemical Space Analysis.
Novel molecular similarity framework that applies Mahalanobis distance metrics to Mol2Vec (Word2Vec-adapted) 300D molecular embeddings. The Mahalanobis angle (θ_M) captures orientation similarity in covariance-transformed descriptor space, identifying analogs that Tanimoto fingerprint similarity misses.
Script developed to build an interactive molecular similarity network to visualize Tanimoto similarity between molecules in a dataset.
ASCEC (Annealing Simulado Con Energía Cuántica) is a Python tool developed by the QFT research group designed to automatically sample molecular configurational PES and perform screening of topological features.
High-performance molecular fingerprint computation and similarity search in Rust. 100% RDKit-accurate MACCS-166, Morgan/ECFP, parallel screening.
ML-driven platform for Glioblastoma drug recommendation using GDSC data. Features multi-model prediction (RF, XGBoost, NN, SVM, KNN), molecular similarity analysis (Tanimoto, MCS, GCN), pathway enrichment, drug interaction checking, and combination therapy optimization with interactive dashboard.
Molecular similarity search tool using RDKit Morgan fingerprints and Tanimoto similarity.
AI-assisted CADD workflows for virtual screening and drug repurposing.
Demonstration of HF:ChemEmbed-v01's Use for Fast Molecular Similarity Search on Large Natural Product SELFIES Dataset
Machine learning framework for Drug-Induced Autoimmunity (DIA) prediction using RDKit molecular descriptors, ECFP fingerprints, Tanimoto similarity graphs, scaffold-aware validation, class-imbalance handling and ensemble classifiers.
Demonstration of ChemFIE-BED's Use for Fast Molecular Similarity Search on Large Natural Product SELFIES Dataset
hacked code to bicluster molecules using rdkit and scikitlearn
To associate your repository with the molecular-similarity topic, visit your repo's landing page and select "manage topics."