Single end read with paired end reads

[tahamasoodi]

Hi,

I've a few samples whose read 2 has not worked and thus contain only read 1. The same samples have been run again and I got around 20x coverage for these samples (read1 & 2). My question, can I merge the these single reads (only read 1) with the paired end reads in order to increase my coverage of samples? If so, will there be any effect on the results?


Thanks,

[pawel.osipowski]

Sorry, I guess my post was off topic here after seeing Brian's answer.

[Brian Bushnell]

I recommend trimming read 2 down to length 1 in the samples that failed. So, read 1 will be full length, read 2 will be length 1, and they will still be paired and correctly ordered; this makes it easy to concatenate the files and map them together. Mappers will ignore length-1 reads.