readcount

Readcount, or read count, is a quantitative measure used in genomics and related fields to indicate how many sequencing reads map to a defined genomic feature or region. It is a fundamental unit in many sequencing-based assays, including RNA sequencing (RNA-Seq), DNA sequencing, and chromatin immunoprecipitation sequencing (ChIP-Seq).

Calculation of read counts involves aligning sequencing reads to a reference genome and then counting reads that overlap a feature such as a gene, exon, promoter, or genomic window. Counting strategies vary in how overlaps are defined (partial versus complete overlap), whether the data are strand-specific, and how reads that map to multiple locations are handled. Common tools for generating read counts include featureCounts, HTSeq-count, and BEDTools' multicov.

Raw read counts are influenced by sequencing depth and feature length, so normalization is required to compare data across samples. Popular normalization methods include counts per million (CPM), reads per kilobase per million mapped reads (RPKM/FPKM), and transcripts per million (TPM). For differential expression analysis and other statistical modeling, count-based methods such as those implemented in DESeq2 and edgeR are routinely used, operating on normalized or raw counts with appropriate dispersion models.

Applications of read counts extend beyond gene expression analysis in RNA-Seq to signal assessment in ChIP-Seq, detection of copy-number variation in DNA-Seq, and other region-based quantifications. Limitations include biases from sample preparation, sequencing technology, and annotation quality, as well as the need for careful experimental design and appropriate statistical models to draw robust conclusions.

In a broader context, readcount can also refer to the number of times a digital content item has been read or viewed, a metric used in web analytics and content management systems, though this usage is separate from sequencing read counts.