Methods Summary: Data were processed with the RGD workflow (GitHub https://github.com/rat-genome-database/RGD_RNAseq_workflows ). Fastq files were downloaded from the Sequence Read Archive and transformed with the sratookit v3.1.1. Raw read quality metrics are generated with FastQC (Babraham Bioinformatics, https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and FastQ-Screen (Babraham Bioinformatics, https://www.bioinformatics.babraham.ac.uk/projects/fastq_screen/). Upon read quality review and species verification, fastq files are aligned to RefSeq genome assembly GRCr8 using STAR version=2.7.10b_alpha_220111 (Alexander Dobin, https://github.com/alexdobin/STAR). STAR alignment files are submitted to RSEM v1.3.1(Bo Li and Colin Dewey, https://github.com/deweylab/RSEM) for expression quantification and output as TPM and counts matrixes. RefSeq annotation release GCF_036323735.1 (https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/036/323/735/GCF_036323735.1_GRCr8/) was used for gene and transcript annotation. For studies processed with the RGD paired-end workflow version 2.1.0 or later, Computed Sex is called from the expression of sex-specific genes: the ratio R = (Xist TPM + 1) / (median TPM of Uty, Ddx3y, Kdm5d and Eif2s3y + 1) is used, with R >= 10 called female, R <= 1 called male, and other values reported as Undetermined; the X:Y chromosome coverage ratio is reported for reference. Library strandedness is inferred for each sample from the proportion of STAR gene counts on the sense and antisense strands, and library preparation (poly(A) selection vs. rRNA depletion) from the number of reads on non-polyadenylated small RNA genes per million gene-assigned reads (<= 350 poly(A) selection, >= 700 rRNA depletion, values between reported as ambiguous). Alignment metrics supporting these calls (fraction of mRNA, intronic, intergenic and ribosomal bases) are computed with Picard CollectRnaSeqMetrics v2.25.0 (Broad Institute, https://broadinstitute.github.io/picard/). Per-sample results are provided in _computed_sample_attributes_GRCr8.txt. MultiQC v1.18 (Seqera, https://seqera.io/multiqc/)is used to generate a comprehensive post-processing quality report. **Disclaimer** Studies submitted to GEO represent a variety of experimental methods, library preparation methods, and processing softwares and algorithms. The HPC RGD workflow is currently limited to transcriptomic studies run on the illumina sequencing platform, with minimal checks on quality. Therefore, the user must be aware of the following limitations to the provided data: The data generated by the HPC RGD workflow may not match the published data due to the variety of data processing methods (software, analysis settings, read trimming, etc.) applied by the submitters to generate the data for publication. Users are encouraged to view the original data by navigating to the GEO accession display tool and viewing the processed data files supplied to GEO by the submitter. All TPM data presented for expression studies by RGD have been processed by the same pipeline, RGD does not attempt to correct for any potential bias or confounding factors. A study may be missing values for one or many samples. This could be a result of technical reasons such as corrupted or missing raw read files or low genome assembly alignment rates. Samples with less than a 50% STAR alignment rate to GRCr8 will not be processed with RSEM to generate abundance estimates. A study may be missing values for one or many genes in the database gene expression table display, RGD does not store TPM values equal to 0. All genes used in the alignment process are output in the matrix files provided in the downloads directory. Computed sample sex may disagree with sex submitted to GEO. For studies processed with the paired-end workflow version 2.1.0 or later, sex is called from the expression of a limited list of sex-specific genes (earlier studies used the X:Y coverage ratio together with these genes). Expression values may be affected by a variety of technical and biological factors, and samples without a clear signal are reported as Undetermined. Where there are significant discrepancies at the study level, RGD curators reach out to data submitters for further clarification. Library strandedness and library preparation method are inferred from the sequence data, not taken from the submitter's description, and may differ from it. These values are currently provided for studies processed with the paired-end workflow version 2.1.0 or later. Samples whose values fall between the calibrated thresholds are reported as ambiguous, and samples with too few reads are reported as unknown. Library preparation methods other than poly(A) selection and rRNA depletion (e.g., total RNA without depletion, or targeted capture) were not part of the calibration and may be called as either method.