Data supporting: Flavour and parthenocarpy evolved independently during fig domestication

# README: Data supporting "Flavour and parthenocarpy evolved independently during fig domestication" ## General information **Corresponding author:** Hidetoshi Ikegami **ORCID:** 0009-0008-7388-7892 **Manuscript preprint:** bioRxiv, MS ID BIORXIV/2026/743143 This dataset accompanies a study integrating whole-genome resequencing, transcriptomics, volatile and metabolite profiling, and taste-sensor data across a diverse panel of fig (*Ficus carica*) accessions, using Bayesian GWAS/TWAS and multi-omics factor analysis (MOFA+) to dissect the genetic architecture of flavour and parthenocarpy. ## Description of files ### Data file S1: Accession panel composition and phenotype metadata - **Format:** .xlsx - **Rows:** 500 (one row per accession) - **Contents:** Accession ID, horticultural classification (common fig / Caprifig / San Pedro / unassigned), parthenocarpy score, sampling location, sequencing status (resequenced / reference), and associated phenotype summary metadata. ### Data file S2: Candidate loci for aroma, skin colour, and anthocyanin traits - **Format:** .xlsx - **Rows:** 258 candidate loci - **Contents:** GWAS/TWAS results for volatile aroma compounds, fruit skin colour, and anthocyanin-related traits. Columns include chromosome, position, marker/gene ID, nearest gene, trait association, effect size, and significance (p-value / FDR-adjusted q-value). ### Data file S3: Functional annotation - **Format:** .xlsx (~47.3 MB) - **Contents:** BLAST, KEGG, and GO functional annotations for genes referenced across the candidate-loci tables (Data files S2 and S4), used for downstream interpretation of candidate gene function. ### Data file S4: MYB101-like and control gene expression under hormone treatment - **Format:** .xlsx - **Rows:** 241 - **Contents:** RNA-seq-derived expression values for *FcMYB101-like* and comparator genes under GA₃ and cytokinin treatment, used to support the functional characterisation of the *Eden* locus candidate gene. ### Data file S5: Normalised expression matrix (TMM) - **Format:** .tsv - **Dimensions:** 45,901 genes (rows) × 142 columns - **Contents:** Trimmed Mean of M-values (TMM)-normalised expression counts across 111 accessions/samples with available RNA-seq data (the RNA-seq participation subset defined in the manuscript's Table S1). This matrix was used as the transcriptome input layer for MOFA+ multi-omics integration and transcriptome-wide association analyses (TWAS) described in the main text. ## Methodological notes - Whole-genome resequencing and variant calling procedures, GWAS/TWAS model specifications, and MOFA+ analysis parameters are described in full in the Methods section of the associated manuscript. - Raw sequencing reads (whole-genome resequencing and RNA-seq) are deposited separately at the DDBJ Sequence Read Archive (accession numbers DRA016899, DRA016900, DRA016910; RNA-seq BioProject PRJDB42532) and are not included in this Dryad deposit. ## Software / tools required - Standard spreadsheet software (Excel, LibreOffice Calc, or equivalent) for .xlsx files. - A text editor, spreadsheet program, or scripting environment (R, Python/pandas) capable of handling tab-separated files for Data file S5, given its size. ## Abbreviations - GWAS: Genome-wide association study - TWAS: Transcriptome-wide association study - MOFA+: Multi-Omics Factor Analysis (version 2) - TMM: Trimmed Mean of M-values (normalisation method) - GA₃: Gibberellic acid - FDR: False discovery rate