Software tools and resources I've developed for structural biology, bioinformatics, and machine learning.
A deep learning framework that predicts glycosyltransferase donor substrates by combining structure-aware protein language models with molecular representations. Utilizes advanced neural architectures to understand both protein sequences and 3D structures.
A Python library for large-scale structural biology analysis that processes thousands of protein structures efficiently using Parquet databases. Enables structural alignment, phylogenetic analysis, binding pocket identification, and physics-based mutation predictions on massive datasets through lazy evaluation.
A proof-of-concept neural network tool for protein sequence similarity searching using deep learning instead of traditional alignment. Encodes proteins into vector embeddings and performs nearest-neighbor searches to detect structural or functional similarities.
A bioinformatics tool for cleaning multiple sequence alignments by removing gap-heavy columns. Supports threshold-based column removal and reference sequence-based gap removal for mutation analysis.
An interactive web database portal for navigating and visualizing evolutionary information within fold A glycosyltransferases. Provides hierarchical classification exploration, sequence alignments, and taxonomic distribution analysis across 9 major clades and 53 families.
A Cytoscape application for modeling how small molecule binders affect signaling pathways. Enables simulation of pathway perturbations with support for multi-step analysis, KEGG pathway integration, and Systems Biology Graphical Notation visualization.