kemalandClaude Opus 5.5
ed4895516d
Graviton3 c7g.8xlarge, 360 GSE165845 arrays: 360/360 GTCs identical between the Rust kernels and the pure-Python paths; 1 worker 78-80 samples/min (pure Python 29.6). Scaling 1->8 workers efficiency 0.87-0.90; above that the 125 MB/s gp3 disk caps the run (~620 samples/min) and, on a RAM disk, a fixed per-worker start-up cost limits a 360-sample batch to efficiency 0.48 at 32 workers (K-025). Worker state is pickled once. docs/validation/P9.md, result JSONs, guide/README/CHANGELOG updates. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01814GKVzZVL3FuF3DeiZXhv
2.8 KiB
SimurgArray
Personal research toolkit for Illumina Infinium genotyping BeadChip data (IDAT → normalization → genotype calling → BAF/LRR → QC → VCF/PLINK; later CN/LOH, PGx, cluster training).
- SIMURGARRAY_MASTER_PLAN.md — architecture, algorithms, roadmap
- SIMURGARRAY_FUNCTIONAL_SPEC.md — 416 itemised requirements (CLI, inputs, outputs, QC, CN, PGx, methylation, edge cases)
docs/validation/— per-phase validation reports ·docs/known_unknowns.md— findings log ·docs/research/— background
Status
v1.0.0 (2026-09-28) — see CHANGELOG.md. IDAT → normalization → GenCall-equivalent genotype calling → GTC v5 →
sample/control QC with an HTML report → plus-strand VCF/BCF → PLINK 1/2, bedGraph, final report; plus a
differential-testing harness. Validation (docs/validation/):
| Check | Result |
|---|---|
2141 vendor GTCs, 1000 Genomes Omni2.5 (--preset autocall-1.6.2) |
normalization bit-identical; GTCs byte-identical in 2138/2141 |
| GSA demo vs gtc2vcf, GTCtoVCF, plink2 | zero unexplained differences |
| 1000 Genomes pedigrees (Omni2.5) | 717/717 parent–offspring pairs; Mendel error median 0.031 % |
| 360 GSA arrays (GSE165845) vs NYGC 30x WGS / GIAB HG001 | concordance 0.9934 / 0.9985; replicates 0.9966 |
Commands: simurg inspect, ingest, normalize, genotype, qc controls|samples|contamination|relatedness|loci|report|suggest-thresholds,
gtc-to-vcf, export plink|bedgraph|finalreport|locus-summary, compare. User guide: docs/user_guide.md;
worked example: examples/gsa_demo.sh. P9 (performance) is done: optional Rust kernels, byte-identical outputs,
~80 GSA samples/min per server core (docs/validation/P9.md). Next (plan §18): P10 packaging, then v2 modules
(cluster training, CN/LOH, PGx).
Quick start
uv venv && uv pip install -e ".[dev]"
uv run pytest tests/unit # no downloads needed
bash scripts/bootstrap_refs.sh # reference open-source repos → refs/ (oracles)
bash scripts/fetch_data.sh p0 # ~1.4 GB of public test data → data/
bash scripts/make_oracle_gtcs.sh # builds gtc2vcf, writes oracle GTCs
bash scripts/make_oracle_qc.sh # locus-summary oracle (BeadArrayFiles, separate venv)
bash scripts/fetch_data.sh p6 # GRCh38 FASTA (~0.9 GB download, 3.1 GB unpacked)
bash scripts/make_oracle_vcf.sh # gtc2vcf oracle VCF for the P6 tests
uv run pytest tests/oracle # real-data acceptance tests (add -m "not slow" to skip the long ones)
uv run simurg inspect data/p0/gsa_v4_pgx/idats_Demo_36/idats_Demo_36/207895980019/207895980019_R03C01_Grn.idat
bash examples/gsa_demo.sh results/gsa_demo 6 # end-to-end run on 6 demo samples
Vendor data files are downloaded for personal research use only and are never committed. See THIRD_PARTY_NOTICES.md.