FASTQ format explained

This post is part 2 of a series on file formats, written for the 2017 UK-KBRIN Essentials of Next Generation Sequencing Workshop at the University of Kentucky. The conference website is hosted here.

#FastQ sequence format

FASTQ was originally developed by the Wellcome Trust Sanger Institute to bind together FASTA sequences with their respective quality data. It is now the standard for high-throughput sequencing output.

The format

FASTQ is a four-line per sequence format. If it looks like the raw sequence of your read takes up more than four lines, you probably have word-wrapping enabled.

@SEQ_ID
GATTTGGGGTTCAAAGCAGTATCGATCAAATAGTAAATCCATTTGTTCAACTCACAGTTT
+
!''*((((***+))%%%++)(%%%%).1***-+*''))**55CCF>>>>>>CCCCCCC65

Line 1

@SEQ_ID is analogous to the >SEQ_ID line of a FASTA file. Like a FASTA title line, everything before the space is the sequence identifier and everything after the first space is the optional sequence description.

Note that different sequence sources have different standards for the description line here. Sequences off of an illumina machine will follow this format:

@Instrument_name:flow_cell_lane:tile_number_in_flowcell:x-coordinate:y-coordinate#index/pair1

So for example,

@myInstrument:1:3:100:200#2/2 describes a read off of myInstrument. The read was on tile number 3 of flowcell lane 1, located at X=100 Y=200, index number 2, pair #2 of a paired-end read.

Note that these standards do change: for example, in Casava 1.8 extra information is added and the description is as follows:

@Instrument_name:runID:flow_cell_ID:flow_cell_lane:tile_number_in_flowcell:x-coordinate:y-coordinate pair:filtered/not_filtered:index_sequence.

If you need to examine the description line of your reads, make sure you know what format is being used.

Line 2

Line 2 is the actual sequence.

Line 3

Line 3 begins with the + character followed by the sequence identifier (and description). This identifeir and description is optional: the + is not.

Line 4

Line 4 contains the quality values for the sequence detailed in line 2. This line must therefore contain the exact same number of characters as line 2: each base has a corresponding sequence score.

Ordered from lowest quality (!) to highest quality (~), these are the quality score characters