P4 validated on 2141 vendor GTCs; legacy AutoCall 1.6.2 preset

- vendor comparison (1000 Genomes Omni2.5, AutoCall 1.6.2.2): every
  computed field identical in 2138/2141 samples; the 3 others trace back
  to the single-locus raw-intensity rule (K-009); sex rule 1 matches all
- AutoCall 1.6.2.x leaves GC10/GC50/#calls/#no-calls at 0 in GTC v3:
  new `--preset autocall-1.6.2` (1.1.2 normalization, sex rule 1, GTC v3,
  legacy statistics); the sample summary keeps the real statistics
- NaN sign canonicalised to the x86 pattern (K-014); report P4.md

README.md

@@ -9,10 +9,10 @@ Personal research toolkit for Illumina Infinium genotyping BeadChip data
 
 ## Status
 
-Phases **P0 (setup)**, **P1 (file parsers)**, **P2 (canonical model, raw X/Y)** and **P3 (normalization)** are
+Phases **P0 (setup)**, **P1 (file parsers)**, **P2 (canonical model, raw X/Y)**, **P3 (normalization)** and **P4 (genotype
-complete — see `docs/validation/`. Each is validated on 2141 vendor GTCs of the 1000 Genomes Omni2.5 set
+calling, GTC output)** are complete — see `docs/validation/`. Each is validated on 2141 vendor GTCs of the 1000 Genomes Omni2.5 set
-(normalization transforms bit-identical).
+(normalization bit-identical; genotypes/scores identical in 2138/2141 samples).
-Available commands: `simurg version`, `simurg inspect <files…> [--json]`, `simurg ingest`, `simurg normalize`.
+Available commands: `simurg version`, `simurg inspect <files…> [--json]`, `simurg ingest`, `simurg normalize`, `simurg genotype`.
 
 ## Quick start
 

docs/known_unknowns.md

@@ -4,6 +4,9 @@ Plan §20 lists the open questions; this file records what the implementation ha
 
 | ID | Date | Topic | Finding | Status |
 |---|---|---|---|---|
+| K-016 | 2026-09-28 | Sex rule of AutoCall 1.6.2.2 | Rule 1 (X heterozygosity) matches all 2141 vendor GTCs; rule 2 (X/Y intensity) differs in 12. | resolved |
+| K-015 | 2026-09-28 | Legacy GTC v3 statistics | AutoCall 1.6.2.2 writes GC10 = GC50 = 0 and #calls = #no-calls = 0 in GTC v3 (it does fill LogR dev and #intensity-only). gtc2vcf's AutoConvert 1.6.3.1 preset fills them, so this is specific to 1.6.2.x. Emulated by `--preset autocall-1.6.2`; the sample summary still reports the real statistics. | resolved; cloud re-run pending |
+| K-014 | 2026-09-28 | NaN sign | NaNs produced by libm (e.g. log2 of a negative extrapolated R) take the platform's default sign: + on aarch64, − on x86-64. SimurgArray always writes 0xFFC00000 (the reference's `-NAN`, x86 behaviour). The aarch64-built oracle differs in 1 locus of 9 GSA samples; oracle tests compare with NaNs canonicalised. | resolved |
 | K-013 | 2026-09-27 | Normalization version of public vendor GTCs | The 1000 Genomes Omni2.5 GTCs (AutoCall 1.6.2.2) are reproduced bit-for-bit by normalization 1.1.2 (109 191/109 191 bins, reserved fields included). | resolved |
 | K-012 | 2026-09-27 | FMA and bit-exactness | gcc on aarch64 contracts a*b+c into FMA by default; the gtc2vcf oracle then differs from the vendor by up to 4 ULP in some bins. Oracles are built with `-ffp-contract=off`. Any future C/Rust port (P9) must also forbid FMA contraction in exact mode. | resolved; constraint for P9 |
 | K-011 | 2026-09-27 | Plan U1 (channel order) | Infinium II X = Red, Y = Grn confirmed on 2141 vendor GTCs (the GTC PDF's "Raw X = green" wording is wrong). | resolved |

docs/validation/P4.md

@@ -0,0 +1,55 @@
+# Phase P4 validation report — genotype calling and GTC output
+
+Date: 2026-09-28. Development and oracle tests on the Pi; vendor comparison on a self-destructing AWS spot instance
+(`run-20260927T233126Z`, 44 min, ≈ $0.19). Summary: `P4_omni2141_summary.json`.
+
+## Delivered
+
+| Item | Module |
+|---|---|
+| Cluster file aligned to manifest order (float32 mean R) | `genotype/clusters.py` |
+| GenCall-equivalent calling: SD flair, six-case half-flattened likelihoods, priors/posteriors, fuzzy sub-scores, score map | `genotype/gencall.py` |
+| BAF / LRR / LogR dev (Peiffer 2006), canonical NaN pattern | `genotype/baf_lrr.py` |
+| Sex estimate, rules v1 and v2 | `genotype/sex.py` |
+| IDAT pair → complete GTC (v5 or v3), TOP-strand base calls, sample stats, presets | `genotype/sample.py` |
+| `simurg genotype` (GTC per sample, `sample_summary.tsv/json` per OUT-060…063, run manifest) | `pipeline/genotype.py` |
+
+## Acceptance (plan §18 P4)
+
+### Against the oracle (gtc2vcf built without FMA), P0 GSA data
+
+| Check | Result |
+|---|---|
+| 36 samples × normalization 1.2.0 and 1.1.2: whole GTC files after aligning the four free-text fields (file names, AutoCall date/version) | **72/72 byte-identical** |
+
+Every data field matches bit for bit: raw X/Y, transforms, genotypes, GenCall scores, TOP base calls, BAF, LRR, call
+rate, sex, LogR dev, GC10/GC50, call counts, percentiles and controls. The only adjustment is the sign of NaNs that
+libm computes (K-014).
+
+### Against 2141 vendor GTCs (1000 Genomes Omni2.5, AutoCall 1.6.2.2, cluster file `HumanOmni2.5-4v1-Multi_B.egt`)
+
+| Field | Samples that differ | Explanation |
+|---|---|---|
+| genotypes | 3 (5 loci in total) | the 3 samples of K-009, whose raw intensity differs in one locus; that difference shifts their normalization |
+| GenCall scores | 3 (18 229 loci) | same 3 samples (the transform change moves every score of the affected bin) |
+| normalization transforms | 2 | same K-009 samples |
+| raw X / raw Y | 1 / 2 | K-009 |
+| call rate, LogR dev | 2 / 3 | follow from the above |
+| base calls | 3 | follow from genotypes |
+| **sex (rule 1)** | **0 of 2141** | rule 2 would differ in 12 samples, so AutoCall 1.6.2.2 uses rule 1 |
+| GC10, GC50, #calls, #no-calls | 2141 | AutoCall 1.6.2.2 leaves these at 0 in GTC v3 (K-015); now emulated by `--preset autocall-1.6.2` |
+
+So in 2138 of 2141 samples every computed field agrees with the vendor; the remaining 3 samples trace back to the
+single-locus raw-intensity rule of K-009. Whole-file byte identity was 0/2141 in this run only because of the four
+legacy-empty statistics; the preset that reproduces them was added after the run (unit-tested, not yet re-run on
+the cloud — expected 2138/2141).
+
+Not covered here: BAF and LRR against vendor output (GTC v3 has no BAF/LRR fields). The 1000 Genomes Sanger set
+ships a GenomeStudio Final Report with BAF/LRR for 177 samples — optional follow-up run (≈ $0.10).
+
+## Performance
+
+| Machine | Result |
+|---|---|
+| Raspberry Pi 5, GSA (683 k loci) | 3.6 s per sample (1 core); 36 samples in 2 min with 4 workers |
+| AWS m7g.4xlarge, Omni2.5 (2.45 M loci), 16 workers | 2141 samples genotyped and compared in 19.6 min; peak RSS 3.2 GB |

docs/validation/P4_omni2141_summary.json

@@ -0,0 +1,615 @@
+{
+  "n_gtc": 2141,
+  "n_compared": 2141,
+  "errors": [],
+  "byte_identical": 0,
+  "samples_with_no_field_diff": 0,
+  "samples_with_field_diff": {
+    "gc10": 2141,
+    "gc50": 2141,
+    "num_calls": 2141,
+    "num_no_calls": 2141,
+    "slide_identifier": 2141,
+    "raw_x": 1,
+    "genotypes": 3,
+    "genotype_scores": 3,
+    "base_calls": 3,
+    "call_rate": 2,
+    "logr_dev": 3,
+    "raw_y": 2,
+    "normalization_transforms": 2
+  },
+  "genotype_diff_loci_total": 5,
+  "score_diff_loci_total": 18229,
+  "max_score_abs_diff": 0.9426311254501343,
+  "cluster_files": [
+    "HumanOmni2.5-4v1-Multi_B.egt"
+  ],
+  "sex_agreement_by_version": {
+    "1": 2141,
+    "2": 2129
+  },
+  "ref_sex_counts": {
+    "M": 1054,
+    "F": 1087,
+    "U": 0
+  },
+  "examples": [
+    {
+      "gtc": "/work/omni/data/5426511011/5426511011_R01C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6847535967826843",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2433210",
+          "0"
+        ],
+        "num_no_calls": [
+          "16790",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511011'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Wednesday, August 25, 2010 7:11:03 PM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511011/5426511011_R02C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6842713356018066",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2436202",
+          "0"
+        ],
+        "num_no_calls": [
+          "13798",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511011'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Wednesday, August 25, 2010 7:24:37 PM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511011/5426511011_R03C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6842615604400635",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435069",
+          "0"
+        ],
+        "num_no_calls": [
+          "14931",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511011'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Wednesday, August 25, 2010 7:37:41 PM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511011/5426511011_R04C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6863988637924194",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497668504714966",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435856",
+          "0"
+        ],
+        "num_no_calls": [
+          "14144",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511011'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Wednesday, August 25, 2010 7:45:50 PM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511013/5426511013_R01C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6843520402908325",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2434354",
+          "0"
+        ],
+        "num_no_calls": [
+          "15646",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511013'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 12:08:26 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511013/5426511013_R02C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.685818076133728",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497965335845947",
+          "0.0"
+        ],
+        "num_calls": [
+          "2439588",
+          "0"
+        ],
+        "num_no_calls": [
+          "10412",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511013'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 12:22:41 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511013/5426511013_R03C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6856454014778137",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497673273086548",
+          "0.0"
+        ],
+        "num_calls": [
+          "2442131",
+          "0"
+        ],
+        "num_no_calls": [
+          "7869",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511013'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 12:37:24 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511013/5426511013_R04C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6856228113174438",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2440841",
+          "0"
+        ],
+        "num_no_calls": [
+          "9159",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511013'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 12:46:33 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511028/5426511028_R01C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6833689212799072",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8496665954589844",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435388",
+          "0"
+        ],
+        "num_no_calls": [
+          "14612",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511028'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 8:13:23 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511028/5426511028_R02C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6857903599739075",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435889",
+          "0"
+        ],
+        "num_no_calls": [
+          "14111",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511028'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 8:28:43 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511028/5426511028_R03C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6847374439239502",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497611880302429",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435026",
+          "0"
+        ],
+        "num_no_calls": [
+          "14974",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511028'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 8:40:16 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511028/5426511028_R04C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6849170923233032",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497409820556641",
+          "0.0"
+        ],
+        "num_calls": [
+          "2433211",
+          "0"
+        ],
+        "num_no_calls": [
+          "16789",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511028'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 8:50:08 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511029/5426511029_R01C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.683781087398529",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497233390808105",
+          "0.0"
+        ],
+        "num_calls": [
+          "2439124",
+          "0"
+        ],
+        "num_no_calls": [
+          "10876",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511029'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 9:12:02 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511029/5426511029_R02C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6832002401351929",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8497028946876526",
+          "0.0"
+        ],
+        "num_calls": [
+          "2437630",
+          "0"
+        ],
+        "num_no_calls": [
+          "12370",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511029'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "M",
+        "2": "M"
+      },
+      "sex_ref": "M",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 9:26:14 AM"
+      }
+    },
+    {
+      "gtc": "/work/omni/data/5426511029/5426511029_R03C01.gtc",
+      "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+      "field_diffs": {
+        "gc10": [
+          "0.6829559803009033",
+          "0.0"
+        ],
+        "gc50": [
+          "0.8496484756469727",
+          "0.0"
+        ],
+        "num_calls": [
+          "2435690",
+          "0"
+        ],
+        "num_no_calls": [
+          "14310",
+          "0"
+        ],
+        "slide_identifier": [
+          "'5426511029'",
+          "''"
+        ]
+      },
+      "sex_by_version": {
+        "1": "F",
+        "2": "F"
+      },
+      "sex_ref": "F",
+      "byte_identical": false,
+      "ref_layout_canonical": true,
+      "metadata": {
+        "autocall_version": "1.6.2.2",
+        "cluster_file": "HumanOmni2.5-4v1-Multi_B.egt",
+        "imaging_date": "Thursday, August 26, 2010 9:40:16 AM"
+      }
+    }
+  ],
+  "params": {
+    "norm_version": "1.1.2",
+    "sex_version": 1
+  },
+  "seconds": 1174.6,
+  "passed": false
+}

scripts/validate_p4_large.py

@@ -37,8 +37,8 @@ ALIGN = (
 _W: dict[str, Any] = {}
 
 
-def _init(bpm: str, egt_dir: str, norm: str, sex_version: int) -> None:
+def _init(bpm: str, egt_dir: str, norm: str, sex_version: int, preset: str) -> None:
-    _W.update(bpm=bpm, egt_dir=egt_dir, norm=norm, sex_version=sex_version, ctx={})
+    _W.update(bpm=bpm, egt_dir=egt_dir, norm=norm, sex_version=sex_version, preset=preset, ctx={})
 
 
 def _ctx(cluster_field: str) -> Any:
@@ -69,7 +69,7 @@ def _canon(a: np.ndarray) -> np.ndarray:  # type: ignore[type-arg]
 def check(args: tuple[str, str, str]) -> dict[str, Any]:
     from simurg.formats.gtc import encode_gtc, read_gtc
     from simurg.formats.idat import read_idat
-    from simurg.genotype.sample import GenotypeParams, genotype_sample
+    from simurg.genotype.sample import genotype_sample, params_for_preset
     from simurg.genotype.sex import SexParams
 
     gtc, grn, red = args
@@ -77,8 +77,12 @@ def check(args: tuple[str, str, str]) -> dict[str, Any]:
     try:
         ref = read_gtc(gtc, check_write_complete=False)
         ctx = _ctx(ref.cluster_file)
-        params = GenotypeParams(
+        params = (
-            norm_version=_W["norm"], sex=SexParams(version=_W["sex_version"]), gtc_version=ref.version
+            params_for_preset(_W["preset"], gtc_version=ref.version)
+            if _W["preset"]
+            else params_for_preset(
+                "current", norm_version=_W["norm"], sex=SexParams(version=_W["sex_version"]), gtc_version=ref.version
+            )
         )
         g = genotype_sample(read_idat(grn), read_idat(red), ctx, params)
         res["cluster_file"] = ctx.cluster_file_name
@@ -110,12 +114,13 @@ def check(args: tuple[str, str, str]) -> dict[str, Any]:
             "ploidy",
             "ploidy_type",
             "dx",
-            "slide_identifier",
         ):
             a, b = getattr(g, k), getattr(ref, k)
             same = a == b or (isinstance(a, float) and np.isnan(a) and np.isnan(b))
             if not same:
                 diffs[k] = [repr(a), repr(b)]
+        if ref.version >= 5 and g.slide_identifier != ref.slide_identifier:  # field 1016 exists only in v5
+            diffs["slide_identifier"] = [g.slide_identifier, ref.slide_identifier]
         if "genotype_scores" in diffs:
             d = np.abs(g.genotype_scores.astype(np.float64) - ref.genotype_scores.astype(np.float64))
             res["max_score_abs_diff"] = float(np.nanmax(d))
@@ -155,6 +160,7 @@ def main() -> int:
     ap.add_argument("--out", required=True, type=Path)
     ap.add_argument("--norm-version", default="1.1.2")
     ap.add_argument("--sex-version", type=int, default=1)
+    ap.add_argument("--preset", default="autocall-1.6.2", help="genotype preset; '' = use --norm/--sex-version")
     ap.add_argument("--workers", type=int, default=os.cpu_count() or 1)
     ap.add_argument("--limit", type=int, default=0)
     a = ap.parse_args()
@@ -184,7 +190,7 @@ def main() -> int:
         a.workers,
         mp_context=mp.get_context("spawn"),
         initializer=_init,
-        initargs=(str(a.bpm), str(a.egt_dir), a.norm_version, a.sex_version),
+        initargs=(str(a.bpm), str(a.egt_dir), a.norm_version, a.sex_version, a.preset),
     ) as ex:
         results = list(ex.map(check, triples, chunksize=2))
 
@@ -218,7 +224,7 @@ def main() -> int:
         },
         "ref_sex_counts": {s: sum(r.get("sex_ref") == s for r in results) for s in ("M", "F", "U")},
         "examples": [r for r in results if r.get("field_diffs")][:15],
-        "params": {"norm_version": a.norm_version, "sex_version": a.sex_version},
+        "params": {"preset": a.preset, "norm_version": a.norm_version, "sex_version": a.sex_version},
         "seconds": round(time.time() - t0, 1),
     }
     summary["passed"] = summary["byte_identical"] == len(results) > 0

src/simurg/cli/main.py

@@ -284,9 +284,18 @@ def genotype_cmd(
     sheet: Annotated[Path | None, typer.Option("--sheet", help="Sample sheet (CSV).")] = None,
     csv: Annotated[Path | None, typer.Option("--csv", help="CSV manifest, checked against the BPM.")] = None,
     cutoff: Annotated[float, typer.Option("--cutoff", help="GenCall score below which a call is a no-call.")] = 0.15,
-    norm_version: Annotated[str, typer.Option("--norm-version", help="1.2.0 (current) or 1.1.2")] = "1.2.0",
+    preset: Annotated[
-    sex_version: Annotated[int, typer.Option("--sex-version", min=1, max=2, help="Sex estimate rule (1 or 2).")] = 2,
+        str, typer.Option("--preset", help="current (default) or autocall-1.6.2 (legacy vendor behaviour)")
-    gtc_version: Annotated[int, typer.Option("--gtc-version", help="GTC format version (3 or 5).")] = 5,
+    ] = "current",
+    norm_version: Annotated[
+        str | None, typer.Option("--norm-version", help="1.2.0 or 1.1.2 (default: from --preset)")
+    ] = None,
+    sex_version: Annotated[
+        int | None, typer.Option("--sex-version", min=1, max=2, help="Sex estimate rule 1 or 2 (default: preset)")
+    ] = None,
+    gtc_version: Annotated[
+        int | None, typer.Option("--gtc-version", help="GTC format 3 or 5 (default: preset)")
+    ] = None,
     autocall_date: Annotated[
         str | None, typer.Option("--autocall-date", help="Fixed value for GTC field 201 (reproducible output).")
     ] = None,
@@ -303,9 +312,13 @@ def genotype_cmd(
 
     if layout not in ("nested", "flat"):
         raise typer.BadParameter("--layout must be nested or flat")
-    if gtc_version not in (3, 5):
+    from simurg.genotype.sample import PRESETS
+
+    if preset not in PRESETS:
+        raise typer.BadParameter(f"--preset must be one of {sorted(PRESETS)}")
+    if gtc_version not in (None, 3, 5):
         raise typer.BadParameter("--gtc-version must be 3 or 5")
-    if norm_version not in ("1.2.0", "1.1.2"):
+    if norm_version not in (None, "1.2.0", "1.1.2"):
         raise typer.BadParameter("--norm-version must be 1.2.0 or 1.1.2")
     rep = Reporter(
         command="genotype",
@@ -327,6 +340,7 @@ def genotype_cmd(
         layout=layout,
         on_duplicate_idat=on_duplicate_idat,
         cutoff=cutoff,
+        preset=preset,
         norm_version=norm_version,
         sex_version=sex_version,
         gtc_version=gtc_version,

src/simurg/genotype/sample.py

@@ -39,6 +39,27 @@ class GenotypeParams:
     gtc_version: int = 5
     autocall_version: str = f"{PRODUCT_NAME} {__version__}"
     autocall_date: str | None = None  # None: run start time (invariant format, OUT-001)
+    legacy_sample_stats: bool = False  # AutoCall 1.6.2.x left GC10, GC50, #calls and #no-calls at 0 (K-015)
+
+
+PRESETS: dict[str, dict[str, object]] = {
+    # default: current vendor software (normalization 1.2.0 / GenTrain 3, sex rule 2, GTC v5)
+    "current": {},
+    # AutoCall 1.6.2.x as used for the public 1000 Genomes Omni2.5 GTCs (validated on 2141 samples, P4)
+    "autocall-1.6.2": {
+        "norm_version": "1.1.2",
+        "sex": SexParams(version=1),
+        "gtc_version": 3,
+        "legacy_sample_stats": True,
+    },
+}
+
+
+def params_for_preset(name: str, **overrides: object) -> GenotypeParams:
+    if name not in PRESETS:
+        raise ValueError(f"unknown preset {name!r}; choose from {sorted(PRESETS)}")
+    values = {**PRESETS[name], **{k: v for k, v in overrides.items() if v is not None}}
+    return GenotypeParams(**values)  # type: ignore[arg-type]
 
 
 @dataclass
@@ -193,4 +214,23 @@ def genotype_sample(
     g.sample_data["num_calls"] = calls.num_calls
     g.sample_data["num_no_calls"] = m.num_loci - calls.num_intensity_only - calls.num_calls
     g.sample_data["num_intensity_only"] = calls.num_intensity_only
+    if params.legacy_sample_stats:
+        g.gc10 = 0.0
+        g.sample_data["p50gc"] = 0.0
+        g.sample_data["num_calls"] = 0
+        g.sample_data["num_no_calls"] = 0
     return g
+
+
+def summary_stats(g: Gtc, intensity_only: npt.NDArray[np.bool_], cutoff: float) -> dict[str, float | int]:
+    """Sample statistics recomputed from calls and scores (independent of legacy GTC field conventions)."""
+    gs = np.sort(g.genotype_scores[(g.genotype_scores > F32(cutoff)) & ~intensity_only])
+    n_calls = int((g.genotypes > 0).sum())
+    n_io = int(intensity_only.sum())
+    return {
+        "gc10": float(percentile(gs, 10)),
+        "gc50": float(percentile(gs, 50)),
+        "n_calls": n_calls,
+        "n_nocalls": g.num_snps - n_io - n_calls,
+        "n_intensity_only": n_io,
+    }

src/simurg/pipeline/genotype.py

@@ -27,7 +27,14 @@ from simurg.formats.gtc import write_gtc
 from simurg.formats.idat import check_pair, read_idat
 from simurg.genotype.baf_lrr import chrom_core
 from simurg.genotype.clusters import align_clusters
-from simurg.genotype.sample import GenotypeParams, ManifestContext, genotype_sample, manifest_context
+from simurg.genotype.sample import (
+    GenotypeParams,
+    ManifestContext,
+    genotype_sample,
+    manifest_context,
+    params_for_preset,
+    summary_stats,
+)
 from simurg.genotype.sex import SexParams
 from simurg.pipeline.inputs import check_paths, resolve_sample_set, sample_dir, sheet_values
 from simurg.signal.normalization.apply import normalize, polar
@@ -77,9 +84,10 @@ class GenotypeOptions:
     layout: str = "nested"
     on_duplicate_idat: str = "error"
     cutoff: float = 0.15
-    norm_version: str = "1.2.0"
+    preset: str = "current"
-    sex_version: int = 2
+    norm_version: str | None = None  # None: from the preset
-    gtc_version: int = 5
+    sex_version: int | None = None
+    gtc_version: int | None = None
     autocall_date: str | None = None
     assay_control_probes: tuple[str, ...] = ("TGA_Ctrl_5716",)
     argv: list[str] = field(default_factory=list)
@@ -133,11 +141,7 @@ def _process(job: Job) -> dict[str, Any]:
             call_rate=float(g.call_rate),
             logr_dev=float(g.logr_dev),
             sex_est=g.gender_char,
-            gc10=float(g.gc10),
+            **summary_stats(g, ctx.manifest.intensity_only, params.cutoff),
-            gc50=float(g.gc50),
-            n_calls=g.num_calls,
-            n_nocalls=g.num_no_calls,
-            n_intensity_only=g.num_intensity_only,
             p05_x=int(g.percentiles_x[0]),
             p50_x=int(g.percentiles_x[1]),
             p95_x=int(g.percentiles_x[2]),
@@ -261,13 +265,17 @@ def run_genotype(opts: GenotypeOptions, rep: Reporter) -> int:
     auto_1_22 = np.array([chrom_core(c) in {str(i) for i in range(1, 23)} for c in manifest.chroms])
     name_idx = manifest.name_index()
     probes = {p: name_idx[p] for p in opts.assay_control_probes if p in name_idx}
-    params = GenotypeParams(
+    params = params_for_preset(
+        opts.preset,
         cutoff=opts.cutoff,
         norm_version=opts.norm_version,
-        sex=SexParams(version=opts.sex_version),
+        sex=SexParams(version=opts.sex_version) if opts.sex_version else None,
         gtc_version=opts.gtc_version,
         autocall_date=opts.autocall_date,
     )
+    rm.set_option("effective_norm_version", params.norm_version, f"preset:{opts.preset}")
+    rm.set_option("effective_sex_version", params.sex.version, f"preset:{opts.preset}")
+    rm.set_option("effective_gtc_version", params.gtc_version, f"preset:{opts.preset}")
     started = _dt.datetime.now()
 
     workers = max(1, min(opts.threads or (os.cpu_count() or 1), len(jobs) or 1))
@@ -335,7 +343,7 @@ def run_genotype(opts: GenotypeOptions, rep: Reporter) -> int:
         "warnings": rep.warning_counts(),
     }
     rm.add_output(summary_tsv, SUMMARY_SCHEMA)
-    rm.schema_versions["*.gtc"] = f"gtc/{opts.gtc_version}"
+    rm.schema_versions["*.gtc"] = f"gtc/{params.gtc_version}"
     rm.status = "completed"
     rm.write(opts.out / "run_manifest.json")
     rep.info(f"done: {ok}/{len(out_rows)} samples genotyped; summary {summary_tsv}")

tests/unit/test_genotype.py

@@ -233,3 +233,25 @@ def test_cluster_alignment_errors(tmp_path: Path) -> None:
     assert e.value.code == "E_EGT_MANIFEST_MISMATCH"
     c = align_clusters(read_egt(tmp_path / "small.egt"), read_bpm(f["bpm"]), allow_missing=True)
     assert int(c.present.sum()) == 1
+
+
+def test_legacy_preset(tmp_path: Path) -> None:
+    from simurg.cli.main import app
+
+    f = _synthetic_run(tmp_path)
+    out = tmp_path / "legacy"
+    r = CliRunner().invoke(app, ["genotype", "--quiet", "--bpm", str(f["bpm"]), "--egt", str(f["egt"]),
+                                 "--idats", str(f["idats"]), "--out", str(out), "--threads", "1",
+                                 "--preset", "autocall-1.6.2"])
+    assert r.exit_code == 0, r.output
+    gtc = next(out.rglob("*.gtc"))
+    g = read_gtc(gtc)
+    assert g.version == 3 and g.gc10 == 0 and g.num_calls == 0 and g.gc50 == 0  # vendor 1.6.2.x conventions
+    summary = json.loads((out / "sample_summary.json").read_text())
+    s0 = summary["samples"][0]
+    assert s0["n_calls"] > 0 and s0["gc10"] > 0  # the summary still reports the real statistics
+    run = json.loads((out / "run_manifest.json").read_text())
+    assert run["effective_options"]["effective_norm_version"]["value"] == "1.1.2"
+    bad = CliRunner().invoke(app, ["genotype", "--bpm", str(f["bpm"]), "--egt", str(f["egt"]), "--idats",
+                                   str(f["idats"]), "--out", str(tmp_path / "x"), "--preset", "nope"])
+    assert bad.exit_code == 2