PolyGenius

Advanced

Deep dives into PolyGenius architecture, the execution and scoring engines, catalogs, and extensibility.

01

Architecture

Data structures, module design, and persistence

  1. 1Data ArchitectureHow PolyGeniusStudy, the shared backbone, models, genotypes and results are structured
  2. 2Persistence and the BackboneWhat a .pgd file physically is, why every object is rebuilt through its own constructor, and what a round trip does and does not keep
02

Infrastructure

Resources, catalogs, the execution engine, caching, and the scoring engine

  1. 1Resources and CatalogsHow external data — panels, chains, variant spaces, LD blocks, tools — is described, retrieved and reused
  2. 2The Execution EngineHow declared work becomes scheduled tasks, how caching decides what runs, and how to read a run
  3. 3The Scoring EngineHow compute$scores() turns a model library and a genotype file into PLINK2 --score calls, and how its memory argument is resolved
03

Extending PolyGenius

Adding algorithms, sources and modules; how the code is organised

  1. 1Extending PolyGeniusAdding a construction algorithm, a GWAS source, an analysis or a plot — and the contracts each must honour
  2. 2Code Organisation and ContributingHow the source tree is arranged, the conventions that hold across it, and what a change is expected to include
04

Reference

API quick reference and glossary

  1. 1API Quick ReferenceEvery accessor by domain, with what it takes and what it returns
  2. 2Glossary and FAQTerms as PolyGenius uses them, answers to recurring questions, and a reporting checklist