htseqcount

HTSeq-count is a command-line utility that is part of the HTSeq Python framework. It counts sequencing reads that overlap annotated genomic features in a reference genome, producing a table of feature-level counts commonly used for RNA-seq differential expression analysis.

Inputs to HTSeq-count consist of an aligned reads file (BAM or SAM) and a gene annotation file in GTF or GFF format. The reads should be sorted by genomic coordinates. Users specify which feature type to count with options such as --type (for example, gene or exon) and the attribute used to identify features with --idattr (for example, gene_id). Strandedness can be indicated to reflect library preparation (--stranded yes, no, or reverse), and the counting order can be set (for example positional). The counting mode determines how reads overlapping features are assigned, with choices including union, intersection-strict, and intersection-nonempty.

Output from HTSeq-count is a tab-delimited table listing features and their counts. By default, reads that do not overlap any specified feature or that are ambiguously assigned may be handled according to the chosen mode. The resulting counts are typically redirected to a file for downstream analysis.

HTSeq-count is distributed as part of the HTSeq package and is commonly used in RNA-seq workflows to generate gene-level count data for downstream statistical analysis. While it remains a widely used tool, newer counting programs and pipelines have appeared that emphasize speed and scalability for very large datasets. HTSeq-count can be installed via PyPI or Bioconda and is compatible with standard Unix-like computing environments.