sequencing read

A sequencing machine does not hand you a whole chromosome spelled out in one piece. It hands you a vast pile of short, individually decoded stretches — each one the sequence of letters from a single DNA fragment. Each of those individual stretches is called a read. The read is the basic unit, the raw output, of essentially every sequencing technology.

When DNA is sequenced, the sample is broken into fragments, and the machine determines the base sequence of each fragment as best it can. A read is the string of called bases — for example AGCTTAGGCA... — that comes out for one fragment, usually with a quality score attached to each base saying how confident the call is. Reads from short-read platforms are typically 50 to 300 bases long; reads from long-read platforms (like nanopore) can be tens of thousands of bases. Reads can be single-end (one fragment, read from one direction) or paired-end (the same fragment read from both ends, which helps software figure out how the pieces fit together).

Reads are the puzzle pieces of genomics. On their own, a billion short reads are just confetti; the value comes from what you do with them — aligning them to a reference genome to find variants, or assembling them into a new genome by their overlaps. Understanding that the raw data is a heap of short, imperfect reads (not a finished genome) is essential to reading genomic claims honestly: how many reads, how long, how accurate, and how deeply they cover each spot all shape what conclusions are trustworthy.