#39 · Tools & infrastructure
Automated curation and standardization of small-molecule data for predictive modeling
The common thread across these items is upstream preparation of small-molecule and ligand data before modeling or simulation. ToxiVerse provides curated toxicity data for roughly 50,000 chemicals alongside bioprofiling and automated QSAR modeling that requires no coding. LigandExplorer extracts and classifies covalent and noncovalent ligands from PDB complexes using residue graphs and LightGBM. ProPrep addresses the simulation side, providing an interactive workflow manager that automates molecular dynamics preparation in AMBER, including structure download, curation, and redox-site parameterization. What is new is the packaging of these steps into automated, accessible workflows with instructional interfaces rather than new model architectures. The items report no comparative benchmarks, so accuracy and relative performance remain unassessed.
Preprints
ProPrep: An Interactive and Instructional Interface for Proper Protein Preparation with AMBER
ProPrep is an interactive workflow manager automating MD preparation in AMBER, integrating structure download, curation, redox-site parameterization, and…
selscape: A Snakemake Workflow for Investigating Genomic Landscapes of Natural Selection
selscape is a Snakemake workflow automating genome-wide natural selection analysis across populations, from input prep and statistics to annotation and…
PopGenAgent: Tool-Aware, Reproducible, Report-Oriented Workflows for Population Genomics
PopGenAgent is an agentic workflow that structures population-genetic analyses as explicit, revisable multi-step plans with preserved intermediate artefacts,…