ASTRA: A Comprehensive Resource of Stress-induced Transcriptional Activity in Human Cell Lines

differential_expression.tsv: A tab-delimited table reporting differential expression results for all untreated versus treated comparisons. Each row corresponds to a single gene within a specific study comparison and includes the Ensembl ID, gene symbol, log₂ fold change, p-value, and details of the stress condition. These values are not averaged across studies but reflect the original within-study analysis. expression_data.tsv: A tab-delimited table containing TPM-based expression values for all genes across all available samples, stored in long format. Each row corresponds to a single gene–sample pair, with columns for the GSM accession, Ensembl gene ID, and TPM value. Since genes are measured in multiple samples, the same gene ID can occur in many rows. gene_annotations.tsv: Gene annotation information retrieved directly from BioMart, including gene symbol, gene biotype, chromosomal location, gene length, and exon count. harmonized_dataset.tsv: A tab-delimited table containing one row per sample (GSM accession) across all studies. For each sample, it includes the internal record ID, GEO series ID (GSE), PubMed ID (PMID), treatment type, stress factor, and detailed biological context such as cell type, tissue, cell group, genotype, and genomic condition. It also specifies whether the sample is categorized as a control or treated. This table serves as the link between individual samples and the broader study metadata, enabling joins with expression data and differential expression results. sample_metadata.tsv: All available sample-level metadata retrieved from GEO, describing each sample’s experimental context. study_metadata.tsv: All study-level metadata from GEO, summarizing study design, stress type, organism, sample counts, and related publication details.datasets.tsv: A tab-delimited table containing the master index of all studies included in the resource. Each row corresponds to a unique study record and contains key metadata, including an internal record ID, the PubMed ID (PMID), the GEO series accession (GSE ID), the total number of samples, treatment type, and stress factor. The table also specifies which GSM accessions are designated as controls or treatments, along with counts for each. Additional fields describe the biological context, including cell type, tissue, cell group, genotype, and any genomic modifications or conditions. This table serves as the primary reference for linking study-level metadata with sample information and expression data across the resource.Publication: Margaritis Tsifintaris, Dimitris Grigoriadis, Michail Sitmalidis, Panagiotis Repanas, Evdoksia K Lazarashvili, Christina Anastasiadi, Ioannis Kavakiotis, Raphael Sandaltzopoulos, Georgios A Pavlopoulos, Antonis Giannakakis, ASTRA: a comprehensive resource of stress-induced transcriptional activity in human cell lines, Nucleic Acids Research, 2025;, gkaf1174, https://doi.org/10.1093/nar/gkaf1174