{
  "study_series_id": "STUDY-SERIES-VERTEBRATE-ADAPTIVE-IMMUNITY-RAG-VLR-033",
  "validation_set_id": "VAL-SET-033-VERTEBRATE-ADAPTIVE-IMMUNITY-RAG-VLR-24COHORT",
  "title": "Sub-Domain Partitioning of Proto-RAG Recombinases, Lamprey VLR Solenoids, and MHC/TCR Beta-Sandwiches Resolves the Dual Origin of Vertebrate Adaptive Immunity",
  "created_at": "2026-11-04T02:55:00Z",
  "prior_beakers": 5,
  "posterior_beakers": 5,
  "tier_1_exploratory_telemetry": {
    "stage": "Tier 1: Exploratory",
    "cohorts_evaluated": 10,
    "taxa_evaluated": 24,
    "total_sequence_span_kb": 388,
    "posterior_z_score": 4.42,
    "beakers_earned": 4,
    "key_findings": [
      "Unpartitioned RAG1/Proto-RAG alignments suffer from compositional drift in the variable N-terminal non-core ubiquitin-ligase/zinc-dimerization tail (1-383 aa, RCFV = 0.158).",
      "Unpartitioned Lamprey/Hagfish VLR alignments conflate hypervariable LRRV internal concave cassettes (dN/dS = 1.42) with the conserved LRRNT/LRRCT disulfide caps and AID/APOBEC CDA1/CDA2 cytidine deaminases (RCFV = 0.166).",
      "Sub-domain partitioning isolates the DDE RNase-H catalytic core + E962/R848 transposition gate in RAG1 (PDB:6B40, TM = 0.922), the LRRNT/LRRCT disulfide caps + invariant threonine stalk in VLR (PDB:2O6S, TM = 0.916), and the C1-set Ig beta-sandwich + CD8/ITAM clasp in MHC/TCR (PDB:1HSA, TM = 0.938)."
    ]
  },
  "tier_2_blind_validation_telemetry": {
    "stage": "Tier 2: Validation (Dual-Pass Blinded SHA-256 Extrapolation -> Unblinding Audit)",
    "cohorts_evaluated": 24,
    "dev_cohorts": 12,
    "reserve_cohorts": 12,
    "total_sequence_span_mb": 1.12,
    "clean_cohorts_recovered": "20/20 (100.0% biotype & clade concordance)",
    "mean_clean_d_ctx_sd": 0.036,
    "reserve_split_z_score": 5.14,
    "posterior_z_score": 5.21,
    "beakers_earned": 5,
    "anomalies_isolated": [
      {
        "token": "BLIND-IMMUNE-06",
        "split": "DEV-03",
        "repository_label": "Branchiostoma belcheri Putative RAG1-RAG2 V(D)J Recombinase Ortholog",
        "pass1_blinded_extrapolation": "Active Cut-and-Paste Proto-RAG Transposase (5-bp TSD + Uncapped E962/D608 Transposition Gate + TIRs; Lacks RAG2 Acidic Hinge & H3K4me3 PHD Finger)",
        "d_ctx_sd": 5.68,
        "error_class": "Functional Transition Misannotation (Active Transposon vs. Domesticated V(D)J Recombinase)"
      },
      {
        "token": "BLIND-IMMUNE-11",
        "split": "DEV-08",
        "repository_label": "Petromyzon marinus Putative TCR/Ig-Like Variable Antigen Receptor",
        "pass1_blinded_extrapolation": "Somatically Assembled LRR Solenoid VLRB + Unpartitioned Hypervariable LRRV Cassettes (Error Type A: RCFV = 0.166, PDB:2O6S TM = 0.916)",
        "d_ctx_sd": 5.52,
        "error_class": "Error Type A (Hypervariable LRRV Cassette Drift & Convergent Receptor Misassignment)"
      },
      {
        "token": "BLIND-IMMUNE-18",
        "split": "RES-04",
        "repository_label": "Callorhinchus milii Full-Length RAG2 V(D)J Recombinase Subunit",
        "pass1_blinded_extrapolation": "Cropped Draft Gene Model Missing C-Terminal PHD Finger (414-487 aa H3K4me3 Reader) & Acidic Hinge SLiM (Error Type B: |J_ij| Dropped 1.724 -> 0.168)",
        "d_ctx_sd": 5.34,
        "error_class": "Error Type B (Truncated Chromatin-Reader PHD & Acidic Hinge SLiM)"
      },
      {
        "token": "BLIND-IMMUNE-22",
        "split": "RES-10",
        "repository_label": "Homo sapiens Reference MHC Class I HLA-A Functional Antigen-Presenting Contig",
        "pass1_blinded_extrapolation": "Unprocessed MHC Class I Pseudogene HLA-H / HLA-J Paralog (dN/dS = 0.892, Premature Exon-4 Stop & Mutated Cys101-Cys164 Disulfide Bridge)",
        "d_ctx_sd": 6.08,
        "error_class": "Paralog Pseudogene Contamination (HLA Class I Pseudogene Filter)"
      }
    ]
  },
  "tier_3_scientific_synthesis_telemetry": {
    "stage": "Tier 3: Scientific Synthesis",
    "discovery_number": 100,
    "discovery_slug": "vertebrate-adaptive-immunity-rag-vlr-tier3",
    "taxa_evaluated": 64,
    "cohorts_evaluated": 32,
    "total_sequence_span_mb": 1.96,
    "cumulative_protein_families": 19,
    "cumulative_taxa": 388,
    "cumulative_span_mb": 13.86,
    "hypotheses": [
      {
        "id": "H1",
        "name": "Proto-RAG Transposon Domestication into Jawed Vertebrate RAG1/RAG2 Recombinase (PDB: 6B40)",
        "rcfv_unpartitioned": 0.158,
        "rcfv_partitioned": 0.020,
        "rcfv_reduction_pct": 87.3,
        "core_dnds": 0.035,
        "potts_j_ij": 1.724,
        "tm_score": 0.922,
        "ufboot_pct": 99.8,
        "z_score": 5.52,
        "beakers": 5
      },
      {
        "id": "H2",
        "name": "Agnathan VLR Leucine-Rich Repeat Solenoid & AID/APOBEC CDA1/CDA2 Gene Conversion Machinery (PDB: 2O6S)",
        "rcfv_unpartitioned": 0.166,
        "rcfv_partitioned": 0.021,
        "rcfv_reduction_pct": 87.3,
        "core_dnds": 0.039,
        "potts_j_ij": 1.682,
        "tm_score": 0.916,
        "ufboot_pct": 99.6,
        "z_score": 5.34,
        "beakers": 5
      },
      {
        "id": "H3",
        "name": "Gnathostome MHC Class I/II & TCR/Ig C1-Set Beta-Sandwich Co-Evolution (PDB: 1HSA)",
        "rcfv_unpartitioned": 0.161,
        "rcfv_partitioned": 0.021,
        "rcfv_reduction_pct": 87.0,
        "core_dnds": 0.032,
        "potts_j_ij": 1.756,
        "tm_score": 0.938,
        "ufboot_pct": 99.7,
        "z_score": 5.41,
        "beakers": 5
      },
      {
        "id": "H4",
        "name": "Falsification Control on Unpartitioned Hypervariable CDR1-3 / LRRV Cassettes & HLA Pseudogenes",
        "rcfv_unpartitioned": 0.174,
        "rcfv_partitioned": 0.174,
        "rcfv_reduction_pct": 0.0,
        "core_dnds": 1.184,
        "potts_j_ij": 0.172,
        "tm_score": 0.0,
        "ufboot_pct": 36.2,
        "z_score": 1.04,
        "beakers": 1
      }
    ],
    "runie_models": [
      {
        "rank": "1st (Best)",
        "architecture": "Super-BLAST -> Minimizer Synteny -> Plan-7 HMM (Core+SLiM vs. CDR/LRRV/Tail Split) -> 2D Dayhoff-6 Recoding -> 4-Stage BIONJ+NNI ML -> Potts Co-Evolution -> 3D Kabsch TM (PDB*:6B40/2O6S/1HSA)",
        "marginal_ln_l": -214612.8,
        "delta_ln_l": 0.0,
        "rcfv": 0.021,
        "posterior_z": "+5.52 SD (Pos) / +4.83 SD (All)"
      },
      {
        "rank": "2nd (Nearest)",
        "architecture": "Super-BLAST -> Plan-7 HMM (Core+SLiM vs. CDR/LRRV/Tail Split) -> 2D Dayhoff-6 Recoding -> 4-Stage BIONJ+NNI ML -> Potts Co-Evolution (No 3D PDB* Superposition Filter)",
        "marginal_ln_l": -214678.4,
        "delta_ln_l": -65.6,
        "rcfv": 0.024,
        "posterior_z": "+4.68 SD"
      },
      {
        "rank": "3rd (Medium)",
        "architecture": "Super-BLAST -> Gotoh 3-State MSA -> IQ-TREE 2* Standard LG+G4 ML* (Pfam*-Clipped DDE/LRR/Ig Core Only, Error Type B: RAG2 PHD, VLR Stalk & CD8/ITAM SLiMs Clipped)",
        "marginal_ln_l": -214819.5,
        "delta_ln_l": -206.7,
        "rcfv": 0.049,
        "posterior_z": "+3.44 SD"
      },
      {
        "rank": "4th (Rejected)",
        "architecture": "MAFFT* Unpartitioned Whole-Protein Concatenation (Includes RAG1 N-Term Tail, Hypervariable LRRV/CDR Loops & HLA Pseudogenes) -> IQ-TREE 2* WAG+G4 ML* (Error Type A Collapse)",
        "marginal_ln_l": -214968.2,
        "delta_ln_l": -355.4,
        "rcfv": 0.166,
        "posterior_z": "+1.04 SD (0/5 Collapse)"
      }
    ]
  }
}