Inflammatory bowel diseases (IBD) gene expression data for the manuscript entitled "TENTACLES: a consensus Machine Learning tool for robust biomarker discovery in heterogeneous data"

RNA sequencing reads related to IBD intestinal biopsies were obtained from the Sequence Read Archive (SRA). All samples were uniformly processed using a standardized custom pipeline to ensure consistency across datasets. Quality control was performed with fastp (v1.0.1), followed by alignment to the human reference genome GRCh38.p14 (NCBI RefSeq assembly GCF_000001405.40) using STAR (v2.7.10) with default parameters. Read strandedness was assessed with RSeQC (v5.0.1), and transcript quantification was carried out using Rsubread (v2.22.1). Publicly available datasets included in this repository: PRJNA248469, PRJNA565216, PRJNA702434, PRJNA728646, PRJNA985602, PRJNA961704. ibd_datasets.rar – contains gene expression count data and clinical metadata from all included studies. PRJNA248469 – Dataset 1 (d1) (Ensemble-based Signature Discovery Cohort) PRJNA565216 and PRJNA985602 – Dataset 2 (d2) (Cross-cohort Signature Evaluation Cohort) PRJNA702434 – Dataset 3 (d3) (Minimal Signature Identification Cohort) PRJNA728646 and PRJNA961704 - Dataset 4 (d4) (Independent Unsupervised Validation Cohort) ibd_datasets.Rdata - contains all tables in a single .Rdata object. TENTACLES_manuscript_vignette.html - contains a demonstration vignette that shows how to work with TENTACLES. ibd_datasets.Rdata is used for this demonstration. TENTACLES_manuscript_vignette.pdf - contains the same demonstration vignette but in PDF format.*Current version: Revised in response to peer-review feedback.