readpair

Readpair refers to two sequencing reads that originate from opposite ends of the same DNA fragment in paired-end sequencing. In a typical experiment, a single DNA fragment is captured and sequenced from both ends, producing two reads that are collectively called a read pair. Reads are labeled read1 and read2, and they usually face toward each other with an expected insert size corresponding to the length of the fragment between the sequenced ends.

Data for read pairs can be stored in two separate FASTQ files or as interleaved entries in a single file. Each read carries a header that links it to its mate, enabling downstream software to treat them as a pair during quality control, alignment, and analysis.

During analysis, read pairs are mapped to a reference genome in a manner that respects their orientation and insert size. The alignment process often uses the information from the paired ends to resolve ambiguous mappings and to improve accuracy. After alignment, paired reads are expected to have concordant mappings with a plausible insert size; discordant pairs can indicate structural variation or library artifacts. Deduplication and quality filtering are commonly applied to read pairs as part of data cleaning.

Read pairs are central to many sequencing applications, including whole-genome sequencing, transcriptome profiling, and targeted sequencing. Key metrics include the mean insert size and its variance, the proportion of properly paired reads, and the rate of PCR duplicates. Understanding read pairing is essential for correct interpretation of sequencing data and downstream analyses.