Concatenate several lanes of Illumina HiSeq reads quickly

If you have raw Illumina HiSeq reads or MiSeq run across several lanes of data, you may need to concatenate them together at the command line before trimming and merging them together. Raw data from Illumina sequencers generally follows a standard naming convention. The files might look like this:


Where ’16’ is the sample ID, followed by the barcode sequence, ‘L00X’ is the lane number, ‘R1’ means forward reads.

An easy way to script this quickly is as follows:

cat `find . -maxdepth 1 | egrep '16.*R1'` > 16_TAGGCGA_R1.fastq &

There’s a lot going on here, so let’s unpack this briefly.

The backquotes (` ) mean that the output of the command is presented as command-line parameters to the enclosing command.

So in this case the ‘find’ command is listing all file names in the current directory at a depth of 1 (no recursion into lower directories).

Next, this is pipe’d (with |) to egrep which searches the list of filenames for a regular expression that indicates a string starting with 16, followed by any number of any characters, then an ‘R1’. Since the expression matches whole strings, this will find the above files.

These filenames are then passed to ‘cat’ as command line arguments. The concatenated files are then redirected with the greater than (‘>’) to the new fastq file. The ‘&’ indicates that the shell run this in the background.

I hope this is useful. I spent about an hour tracking this all down, so I wouldn’t have to process dozens of samples by hand.