Feasibility of cross-publication SNP-effect synthesis for livestock growth traits: a purposive audit of 35 reports, with a full-corpus accessibility and completeness screening, from a 193-report PubMed corpus

This is the data and supplementary package accompanying: "Feasibility of cross-publication SNP-effect synthesis for livestock growth traits: a purposive audit of 35 reports, with a full-corpus accessibility and completeness screening, from a 193-report PubMed corpus"Albert Iribagiza, Vanessa Nibigira, Hermenegilde Niyongabo, Emmanuel Mvuyekure This is a purposive methodological feasibility audit, not a PRISMA-compliant systematic review. It examines whether SNP-level effect estimates (allele, direction, magnitude, standard error) can be recovered and reused across independent published GWAS reports on growth-related traits in pigs, sheep and goats, rather than estimating biological effect sizes or reporting prevalence. Package contents:- Main_Manuscript.docx — current manuscript text (structured six-heading abstract; accessibility screening integrated into the main methods/results sequence; AI-use declaration included)- Highlights.docx — five-bullet highlights summary- Supplementary_Tables_final.xlsx — full supplementary workbook (Tables S1-S13): report-level corpus characteristics, bibliography, candidate-source and overlap registers, direct-verification log, dated PubMed search documentation, independent human-reviewer screening agreement statistics, the 35/16-report variant-audit selection lists (S12), and a full-corpus accessibility and 6-field completeness audit of all 157 eligible reports with the itemized 36-report exclusion list (S13)- PubMed_2026-08-29_669_records.csv — raw export of the 29 August 2026 PubMed search that identified the 193-report reference corpus- PubMed_2026-08-29_195_corrected_fulltext_queue.csv — corrected full-text eligibility queue (193 included, 2 excluded)- PubMed_2026-09-20_628_search_records.csv — recovery search used to check corpus completeness and flag update candidates- Supplementary_Search_Screening_Reconciliation.docx — verbatim search queries and eligibility-flow reconciliation- README.txt — file-by-file description and scope notes Reference corpus: 193 reports (113 pig, 60 sheep, 20 goat). Variant-level audit: 35 reports matrix-audited, 16 directly inspected (44 unique reports; no goat report assessed at variant level). A separate full-corpus accessibility screening confirmed 36 of the 193 reports (18.7%) as non-open-access; the remaining 157 (81.3%) were audited against a 6-field per-SNP effect-reporting checklist (Supplementary Table S13), with recovery rates ranging from 68.8% for genomic coordinates down to 6.4% for standard errors. Independent duplicate title-level screening of 376 update-candidate records was completed by two non-co-author human reviewers (Supplementary Table S10; percent agreement and Cohen's/Fleiss' kappa reported); a separate three-model AI-ensemble comparison is reported as a post hoc robustness check only (Supplementary Table S11) and does not substitute for the human screening. Keywords: GWAS, genome-wide association study, livestock genetics, growth traits, pig, sheep, goat, SNP effect, meta-analysis feasibility, systematic screening, data reuse License: CC-BY 4.0 Corresponding author: Albert Iribagiza ([e-mail]), ORCID 0009-0004-4417-9265