**Background:** RNA-sequencing (RNA-seq) has become the dominant technique for transcriptional profiling, enabling the study of gene structures, splicing patterns, and expression levels. Large volumes of raw sequencing data have been deposited in public repositories such as NCBI SRA, ENA, and GSA, but accessing and analyzing these data remains challenging for wet-lab and field researchers who lack bioinformatics experience and computational resources. While comprehensive RNA-seq databases exist for other crops (e.g., Arabidopsis, wheat, Brassica napus), barley lacked a well-integrated, up-to-date resource that covers multiple subspecies and the latest reference genomes.
**Methods:** The authors collected 56 studies comprising 3,492 publicly available barley RNA-seq libraries from NCBI SRA. Raw reads were downloaded using SRAtoolkit v2.10.8 and converted to FASTQ format. Quality control was performed with Trimmomatic v0.36 (parameters: Minlen=90, Trailing=3, Leading=3, SlidingWindow=4:5). Reads were aligned to four reference genomes—cultivated barley Morex V2 and V3, wild barley B1K-04-12, and Tibetan hulless barley Zangqing320—using Hisat v2.1.0. SAM files were converted to BAM format and sorted with SAMtools v1.3.1. FPKM values were calculated using Stringtie v1.3.5. Functional annotations (GO, KEGG, PFAM, SMART, EC numbers) were assigned using eggNOG-mapper v2. Orthologs to rice and Arabidopsis thaliana were identified using Inparanoid v8.0. The web interface was built with HTML, JavaScript, and CSS on the front-end, with PHP and MySQL v5.6.50 on the back-end, hosted on a Tencent Cloud server (4 Intel Xeon Platinum 8255C CPUs at 2.50 GHz, 8 GB RAM, CentOS v7.9).
**Key Results:** Over 12 TB of high-quality clean data were generated. Alignment statistics showed that most reads aligned concordantly exactly once to the reference genome, supporting data reliability. For BioProject PRJEB14349 (RNA-seq of 16 developmental stages), the mean pair-wise Pearson correlation coefficient among replicates was 0.8069 (standard error 0.1368), indicating high reproducibility. PCA confirmed strong clustering of replicates from the same stage/tissue and clear separation between different stages/tissues. The database supports queries by gene ID (up to 500 genes per batch) or by sequence similarity via BLAST (BLASTN, BLASTP, TBLASTX, etc.). Expression data are displayed as interactive heatmaps (rendered with Plotly.js) and downloadable in PNG and CSV formats. The database provides comprehensive functional annotations including PFAM, SMART, GO, KEGG, and functional descriptions, as well as links to homologous genes in rice and Arabidopsis.
**Clinical Implications:** BarleyExpDB is a bioinformatics resource rather than a clinical study, so direct clinical implications are not applicable. However, the database has significant implications for agricultural and crop science: it enables researchers to efficiently mine publicly available transcriptome data to identify genes involved in stress tolerance, developmental regulation, and agronomic traits. By supporting four barley genomes (including wild and Tibetan hulless barley), it facilitates comparative genomics and breeding research. The authors plan regular updates (at least twice per year), integration of telomere-to-telomere assemblies, pangenome data, RNA editing analysis, and single-cell/spatial transcriptomics data, which will further enhance its utility for functional genomics and molecular breeding in barley.