Overview#
Automate docking, then reuse the same results for analysis
Prepare structures, build docking inputs, run campaigns across engines, and keep poses, scores, and interactions queryable in one SQLite database.
Follow the workflow
Structure, preprocess, dock, postprocess, database, and automation pages stay grouped under one tutorial hub.
Understand the design
See why ProDock uses a pose-centric database and how the package is split into reusable layers.
Scan the package
Start with a short API page, then move to a larger automodule-based reference under the same API section.
Rerank GNINA and DiffDock poses
Validate paired inputs, optimize descriptor thresholds, replay train-selected thresholds on test data, and interpret generated results.
Record an auditable campaign
Separate source inputs from generated artifacts and capture the configuration, versions, labels, and checks needed to repeat a run.
Store once, analyze many times#
Package map
Core orchestration stays compact while structure, preprocess, docking, postprocess, and database modules remain reusable.
Database overview
Catalog tables feed a central poses table, which branches into score and interaction storage.
