MiSeq FastQ Header for pair-end data

[golharam]

So the Miseq fastq header for pair-end data has changed. I used to see a /1 or /2 at the end of the read name. Now I get something like this for the first read:

@HWI:1:X:2:1101:1094:2143 1:N:0: CGGCTATG-ATAGAGGC

and this for the second read:

@HWI:1:X:2:1101:1094:2143 4:N:0: CGGCTATG-ATAGAGGC

I've seen the the number 3 or 4 in the second read but not sure what they mean, and haven't seen an explanation anywhere. Does anyone know what this number signifies?

[Brian Bushnell]

That's good to know; I had only seen 2 for read 2 - now I need to update my header-detection routines. I imagine it has to do with how the machine counts sequencing phases for dual-index:

1: Read 1
2: Barcode 1
3: Barcode 2
4: Read 2

It's not unusual for Illumina to randomly change stuff like that with software updates, which is pretty annoying.

[GenoMax]

Can you tell us where the data was analyzed, on instrument/in BaseSpace? Was the analysis done using standard parameters (i.e. no separate fastq files for tag reads).

[kcchan]

If the headers are starting with @HWI it's almost certainly generated using BCL2FASTQ and not MiSeq Reporter, which uses @M0xxxx.