CORBIC ROB — data and code bundle (Zenodo)

CORBIC ROB — data and code bundle (Zenodo) Built 2026-10-29 by rob_W04_zenodo_bundle.py from the accepted runs of the ROB project (LAB v5). Nothing derived was edited by hand. Layout runs/<STEP>__<time>-<id>/ one accepted run: run.json (command, host, inputs/outputs md5, exit code), run.log, code/ (exact script that ran),conda_list.txt, out/ (outputs up to the size cap).numbers.py -> numbers.tsv: every number in the manuscript, read from accepted outputs (column file, locator, md5).DECISIONS.md, accept_log.tsv, accepted.tsv: decisions and the acceptance record. workflow/scripts, lib, config, data_entry: code andhand-entered tables.manuscript/: text parts. omitted_files.tsv: output files above the size cap (path, bytes, md5); they are not in this archive. Accepted runsstepB23_T00_artifact_checksumsB23_T00_manifest_checkB23_T01_alignment_namingB23_T01_loo_inputsB23_T01_preflightB23_T02_check_onlyB23_T03_qc_L4b_PGRPB23_T03_qc_L4b_TollB23_T03_qc_pilotB23_T03_qc_reconcileB23_T03_rebuild_DreddB23_T03_rebuild_FADDB23_T03_rebuild_PGRP_OG2B23_T03_rebuild_PGRP_OG3B23_T03_rebuild_RelishB23_T03_rebuild_Toll_OG0B23_T03_rebuild_Toll_OG10B23_T03_rebuild_Toll_OG11B23_T03_rebuild_Toll_OG2B23_T03_rebuild_Toll_OG5B23_T03_rebuild_cactusB23_T03_units_draftrunB23_T00_artifact_checksums__20260917-183228-1bfbB23_T00_manifest_check__20260917-183227-aa8aB23_T01_alignment_naming__20260917-185225-dcf6B23_T01_loo_inputs__20260917-185705-de97B23_T01_preflight__20260917-184603-c66fB23_T02_check_only__20260928-100247-ac76B23_T03_qc_L4b_PGRP__20260928-134801-e6ddB23_T03_qc_L4b_Toll__20260928-134750-167eB23_T03_qc_pilot__20260928-135003-f661B23_T03_qc_reconcile__20260928-135011-2feeB23_T03_rebuild_Dredd__20260928-145603-edf4B23_T03_rebuild_FADD__20260928-151128-3b56B23_T03_rebuild_PGRP_OG2__20260928-151124-b376B23_T03_rebuild_PGRP_OG3__20260928-151126-80b5B23_T03_rebuild_Relish__20260928-151130-6e9cB23_T03_rebuild_Toll_OG0__20260928-151114-53d1B23_T03_rebuild_Toll_OG10__20260928-151120-dc7eB23_T03_rebuild_Toll_OG11__20260928-151122-eb0aB23_T03_rebuild_Toll_OG2__20260928-151116-fef2B23_T03_rebuild_Toll_OG5__20260928-151118-3413B23_T03_rebuild_cactus__20260928-151132-b764note249/249 OK, 0 MD5_DIFF; independent checksum layer overdata/genomes confirms the tree626/627 OK, 0 MD5_DIFF; the single MISSING entry(CORBIC_STATE.md) is present at workflow/scripts/ with themanifest md5 — relocated, not lost; tree identical tnaming resolved by measurement: 21 families units have alnshorter than untrimmed, all 10 pilot units have aln longerthan trim; codon.aln.fa therefore denotesall 10 pilot units: stored LOO reference fit used the untrimmedalignment while the canonical fit used the trimmed one(DIFFERENT_INPUT); all 17 families/reconcG1 passed: busted/relax/gard options inventoried from help;error-sink, starting-points, grid-size, multiple-hits all presentin HyPhy 2.5.100 env hyphy (no hyprun_relax_v3 check-only on one families and one pilot unit;selftests 16/6/3 PASSv3 reproduces stored B21 PGRP sweep (compare_v2: 0differences); 3 TABLE_ONLY: CDS differs from annotatedprotein (OG1 cerana offset; OG3 Bombus CDS covers N-tev3 reproduces stored B21 Toll sweep (compare_v2: 0differences); L4b sensitivity changes no verdictD41/D57 QC of the 10 pilot units; all mapped; new FAILs allL3+L4b onlyD41/D57 QC of cactus, Stat_long, Stat_short; all mapped;cactus H. saltator FAIL on L2+L3D58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 rebuild; full alignment and topology identical to frozen;stored single-start BUSTED fit sits at lower likelihood than thisfit (optimiser, see D60 proposalD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 rebuild; full alignment and topology identical to frozen;this single-start BUSTED fit sits at lower likelihood than thestored fit (optimiser, see D60 propD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsD58 full+postqc rebuild; D59 curation replayed; BUSTEDsettings canonical in both variantsB23_T04_fdrB23_T04_fdr_rescuedB23_T04_seed_checkB23_T05_busted_fitsB23_T05_endpointsB23_T05b_warm_eB23_T05r_rescueB23_T06_topology_fitsB23_T06_topology_tableB23_T07_evalB23_T07_eval_rescuedB23_T07_loo_bustedB23_T08_relax_fdrB23_T08_relax_fitsB23_T09_gardB23_T10_modelsB23_T12_descriptiveB23_T12_tablesB23_T13_gard_segmentsB23_T13_matrixB23_T13_matrix_v2B23_T13_relax_matrixB23_W01_config_snapshotB23_T04_fdr__20260928-194721-9cdbB23_T04_fdr_rescued__20260930-135601-0f3eB23_T04_seed_check__20260928-170057-91cfB23_T05_busted_fits__20260928-180325-3920B23_T05_endpoints__20260929-102946-464eB23_T05b_warm_e__20260929-183650-bf72B23_T05r_rescue__20260930-064028-1b21B23_T06_topology_fits__20260929-192709-8b69B23_T06_topology_table__20260929-124628-dce9B23_T07_eval__20260929-131923-cf18B23_T07_eval_rescued__20260930-135602-6d8fB23_T07_loo_busted__20260929-102257-2b4bB23_T08_relax_fdr__20261005-060213-abe1B23_T08_relax_fits__20261001-113002-10a2B23_T09_gard__20260930-155613-aa3cB23_T10_models__20261001-042545-dd14B23_T12_descriptive__20261006-080857-f84fB23_T12_tables__20261005-131919-5a2eB23_T13_gard_segments__20260930-082551-2c40B23_T13_matrix__20260930-135738-abacB23_T13_matrix_v2__20261001-064806-2db2B23_T13_relax_matrix__20261005-060358-e238B23_T03_units_draft__20260928-132502-c207D62 sets: canonical and postqc both m=28, no member set top=1, two UNSETTLED members (OPT_UNCONFIRMED); fieldcounts consistent; full-precision qD69c: 7 T05 objects overlaid; post-QC q<0.05 unchanged(7/28); unsettled 2->1; regression without –rescue identical toB23_T04_fdrHyPhy 2.5.100: ENV=RANDOM_SEED does not make BUSTEDreproducible (1 and 4 threads); Toll_OG10 full, same settings:fits land in two optima with different calls;D60 fits plain + E_primary: all object-variants settled(CONFIRMED or OPT_UNCONFIRMED), none FAILEDSTOPPED_TIME/NR; every replicate exit 0 and parsed, noemptyE1-E3 and D63 checks on accepted T05/T04: field countsconsistent; D63b changes no call; D63a gives E_BELOW_PLAINfor six nested objects (E call NR on AX2), nonD65: 5 objects warm_start_took yes, D63a OK, E_call NS;MyD88 WARM_NR (plain phase did not reach the T05 best logLin 20 reps; no warm start attempted), D63a E_D69: 40 targets, 1230 new replicates all parsed, none skipped;29 CONFIRMED, 11 OPT_UNCONFIRMED; 6 calls NS->SIG(Toll_OG8S1, pelledel_apis_mellifera, Toll_D66, D66a, D66d: 28 bases = post-QC BH set; 49 fitted + 7SAME_AS_S1 (consistent); 46 CONFIRMED, 3OPT_UNCONFIRMED (MyD88 S2, Toll_OG5__postqc S1+S2); all350 rD66 table: every T05 object; every split labelled; field countsand row counts consistent; inspectedD67 + D67a; deletions.tsv and D40 verdicts identical to theunaccepted first run B23_T07_eval__20260929-131325-066f;field counts consistent; inspectedD69c: 36 rescued objects used; D40 prune PASS 2->3(PGRP_OG1); regression without –rescue identical toB23_T07_evalD64: every object settled (CONFIRMED orOPT_UNCONFIRMED), none FAILED/STOPPED_TIME/NR; allreplicates parsed; every base of sets A and B present; fieldcounts cD73/D73a: m=26 P members, 2 NOT_TESTABLE listed, no p setto 1, 2 unsettled; built on the accepted B23_T08_relax_fits;inspected 2026-10-05D72/D72a: 36 fitted objects x 40 replicates, exit 0, no replicateskipped, no FAILED; 14 JOB_FAIL replicates count as non-hits(D72); inspected 2026-10-05D70: GARD OK on Toll_OG3 (4 segments) and Toll_OG6 (2); 4segments >= 100 codons, IQ-TREE exit 0, 40 replicates each,all 160 parsed, all CONFIRMEDD71: PGRP_OG1 isoforms FAIL (used X1, best X3), Toll_OG3domains PASS (no TIR in any member incl. reference),Toll_OG6 domains PASSW01: RQ4 taxon-level test (D75); sums agree with QC andrebuildsT12 tabulations (values, crosstab, numeric) of acceptedoutputs; field counts consistent; selftest passed; read bynumbers.py v2D68a: PGRP_OG0 and PGRP_OG1 applicable (segments >= 100codons fitted as _p2); PGRP_OG2, PGRP_OG3GARD_ON_OTHER_ALIGNMENT (AX7 NR until T09)D68, D68a, D69c: 28 P rows, 13 C rows; sources = acceptedrescued runs; verdicts consistent with conditions; inspectedD68, D70: T09 segments and calls added; differs fromB23_T13_matrix only in AX7 and verdicts of Toll_OG3, Toll_OG6(both NOT_ROBUST:AX7); canonical BUSTED matriD73/D73a: 26 rows, verdicts NOT_ROBUST:AX8 orUNDETERMINED:AX3,AX5 D56 units table draft; no ERROR, (AX5 NR no by uncovered design); built treefile; on everyaccepted stored BUSTED/RELAX T08 fits and fdr; fit read inspected the trimmed 2026-10-05 alignment; CHECK:no orthoset for pilot units and cactusB23_W02_fig2B23_W02_fig3B23_W02_fig4B23_W02_fig5B23_W02_refcheckB23_W03_fig1B23_W03_fig1_selftestB23_W03_hmmer_versionB23_W03_hmmer_version_selftestB23_W03_refcheckB23_W03_refcheck_selftestB23_W04_fig1B23_W04_fig1_selftestB23_W04_fig2B23_W04_fig2_selftestB23_W04_fig4B23_W04_fig4_selftestB23_W04_refcheckB23_W04_refcheck_selftestB23_W04_stablesB23_W04_stables_selftestB23_W02_fig2__20261024-093819-c933B23_W02_fig3__20261006-135703-36f6B23_W02_fig4__20261006-112939-6ee9B23_W02_fig5__20261006-153047-d795B23_W02_refcheck__20261006-090856-29dbB23_W03_fig1__20261025-114908-4b77B23_W03_fig1_selftest__20261025-114904-e9e0B23_W03_hmmer_version__20261029-102850-d4fcB23_W03_hmmer_version_selftest__20261029-102850-6297B23_W03_refcheck__20261029-095751-a5f1B23_W03_refcheck_selftest__20261029-095750-259eB23_W04_fig1__20261029-170302-c67dB23_W04_fig1_selftest__20261029-170248-17a7B23_W04_fig2__20261029-170306-d84fB23_W04_fig2_selftest__20261029-170253-a039B23_W04_fig4__20261029-170313-4b91B23_W04_fig4_selftest__20261029-170258-af81B23_W04_refcheck__20261029-150457-3b41B23_W04_refcheck_selftest__20261029-150456-24d4B23_W04_stables__20261029-171333-ce5dB23_W04_stables_selftest__20261029-171330-2973F2 artefact anatomy and paired experiment; all checks yes (8numbers.tsv keys, rq1 vs deletions dLRT cross-check); withoutmodel LRT from FAIL_single deletion oF3 robustness matrices and attrition; 14 numbers.tsv keys and10 attrition stages match; verdicts equal first FAIL in D40order; figure inspected row by rowF4 optimiser figure; all 8 count checks vs numbers.tsv yes;panel b calls at p level as AX6; rescue labels match rule (SIGonly if confirmed and LRT above 2ln10F5 input quality; 63 checks yes incl. threshold consistency onevery QC row; figure inspected (panel b equals Table 2)43 DOI found, 0 fetch errors, 120 fields match; only diffWeisman 2022 pages .e2 (Cell Press format); query rowsinspected: Verdonk 10.64898/2026.06.02.729680,F1 pipeline schematic; 38 checks yes against Table 1 W03a; nodata; figure inspectedF1 selftest: all cases pass, exit codes 0 and 4 as designedhmmscan version read from binary of accepted domain scanselftest hmmer version54 refs, one known pages diff weisman2022selftest refcheckF1 pipeline schematic, J Mol Evol format (174 mm, 8 ptLiberation Sans, EPS); 38 checks yes against Table 1 W03a; nodata; figure inspectedF1 v3 selftest: all cases pass, exit codes 0 and 4 as designed;Liberation Sans bundledF2 artefact anatomy and paired experiment, W04 format (174mm, 8 pt, EPS); all checks yes (8 numbers.tsv keys, deletionscross-check); canonical Toll_OG9 differF2 v3 selftest: all cases pass; exit codes 0 and 4 as designedF4 optimiser figure, W04 format (174 mm, 8 pt, EPS); all 8count checks vs numbers.tsv yes; figure inspectedF4 v3 selftest: all cases pass; exit codes 0 and 4 as designed55 refs, 0 fetch errors, 152 fields match; one known pages diffweisman2022 (.e2); bachler, iqtree3, verdonk, pfam rows nowwith DOI; redelings2014 Crossref 31:selftest refcheckSupplementary Tables S1-S10 (Online Resource 2): 34 sheetsfrom accepted runs, all checks yes (row counts, numbers.tsv:qc 254, relax 27, rescue 40, GARD 6); woS1-S10 xlsx selftest: all cases pass, exit codes 0 and 4 asdesigned Figures and tables -> run Figures F1-F5: steps B23_W03_fig1, B23_W02_fig2, B23_W02_fig3, B23_W02_fig4, B23_W02_fig5 (scripts in workflow/scripts/rob_W03_fig1.py,rob_W02_fig2.py … rob_W02_fig5.py). Supplementary Tables S1-S10 (Online Resource 2): the accepted run named in accepted.tsv for thesupplementary-tables step, if present