SeqrunA first look at a FASTQ file
FASTQ is the raw output of short-read sequencing: one record per read, four lines each. A few minutes reading it before you commit to a pipeline catches the obvious problems early.
Four lines per read
- Line 1 — header: starts with
@and names the read (instrument, lane, coordinates, and the mate number for paired-end). - Line 2 — sequence: the base calls, A/C/G/T (and N for an ambiguous call).
- Line 3 — separator: a single
+(optionally repeating the header). - Line 4 — quality: one character per base, encoding the Phred score.
How quality is encoded
Modern Illumina data uses Phred+33: each quality character’s ASCII code minus 33 is the Phred score. So the character I (ASCII 73) is Q = 73 − 33 = Q40, and # (ASCII 35) is Q2. The quality string lines up base-for-base with the sequence, so position 10 of the quality line is the confidence in base 10.
What to check first
- Read count and length: do they match what you ordered? A truncated or half-size file shows up here immediately.
- Quality along the read: a gentle drop toward the 3′ end is normal; a cliff is not.
- GC content: a value far from what your organism or library predicts can flag contamination or bias.
- Over-represented sequences: usually adapter read-through — worth identifying and trimming.
None of this needs a heavyweight pipeline: aggregate read count, length, per-base quality and GC from the file itself and you have your first, honest picture of the run.