Data and code for: Juvenile protein restriction compresses sexual dimorphism in mice

This deposit contains the curated wet-lab measurements, the R analysis pipeline, and the supplementary-data build outputs supporting the manuscript "Juvenile protein restriction compresses sexual dimorphism in mice" (Joly, Rebiffé, et al., 2026). It supersedes version 3, which described an earlier version of the study under a different title; 83% of the analysis code has changed since, and four raw files carry a mouse-identifier reassignment documented in their codebooks. Contents (305 files): data/raw/ — 153 files: 71 curated CSVs (body weight, naso-anal length, femur length, IGF-1, FGF21, leptin, OGTT/ITT/PTT, insulin, hepatic qPCR, liver metabolomics, serum amino acids, transcriptome inputs, reproductive maturation, estrous cytology, three reproductive-output trials, orchidectomy cohort, serum steroid panel) each with a companion *_codebook.md. Includes four Oil Red O representative micrographs (JPEG, 2560 × 1920, one per Sex × Diet group) and the source acquisitions of the MUP western blot. data/metadata/, data/gene_sets/, data/transcriptome_counts/, data/transcriptome_counts_green/, data/external/, data/processed/ — transcriptome sample metadata, 15 curated gene sets (ATF4 and PPARα targets; STAT5b, BCL6, GH-axis dimorphism and hepatic feminization modules) with their own README, our count matrices, the published adult count matrix re-analysed for the cross-age comparison (GSE181301), and one external supplementary table from Rampersaud et al. (eLife RP91367) redistributed under CC-BY-4.0 with its source declared. scripts/ — 70 numbered analysis scripts, 13 helper modules, 4 Data S1–S4 builders and one diagnostic script. run_all.R is a single-entry-point runner that rebuilds every deposited result from the raw inputs; our last complete run of the full project tree executed 87 steps in 11.1 minutes with no failure. Note: figure assembly (scripts/figures/) and the assembled figures (results/figures/) are deliberately excluded from this deposit, so run_all.R completes the 69 analysis steps and the 4 Data S builders and then stops when it reaches the figure scripts. This is expected and is stated in README.md. results/supplementary/Data_S1–S4.xlsx — hepatic metabolomics and serum amino acids (S1), hepatic transcriptomics (S2), orchidectomy qPCR (S3), mouse metadata, cohorts, statistical methods and per-figure sample sizes (S4). environment/sessionInfo.txt — captured R 4.4.3 session (aarch64-apple-darwin20, macOS 27.0) with exact package versions, recorded at the end of the run that produced the deposited results. Root — README.md, LICENSE, CITATION.cff, plus .here, Dimorphism_v2.Rproj and init_project.sh so that the project root resolves wherever the archive is unpacked. Cohorts covered: nine. C57BL/6N wild-type factorial Sex × Diet (AIN-93G control, 18.5% kcal protein, versus a matched-lipid protein-diluted variant, 4.8% kcal protein, from P21 to P56); prepubertal orchidectomy crossed with diet, with an LC-MS/MS serum steroid panel; hepatocyte-specific Fgf21 knockout on a C57BL/6J background; and three reproductive-output cohorts (male non-competitive B6N × OF1 mating over three cycles, female three-arm CD/LPD/Switch periconceptional design over two parities, and female dietary rescue at P70 paired with proven studs). Raw sequencing data (FASTQ) are in the Gene Expression Omnibus under accession GSE302336. Companion preprint: bioRxiv 10.1101/2025.07.09.663702