isolate reads from a paired-ended fastq files

[lorenzo.C]

hi folks,
I have pair ended fastq files (let's call them 01_R1 and 01_R2). I need to make a new pair of fastq files that contains just the reads that fulfill this requirement: the reads of R2 should be within a certain range (let's say no more than 100bp far) from the respective read in R1.
any suggestion on how to approach this?

[bluesmartini]

Hello,

I would say that most of the aligners have this option. Which aligner do you want to use?

[lorenzo.C]

BOWTIE2 would be great! thanks for the help!

[bluesmartini]

with Bowtie2, the (very extensive) documentation mentions the following options :

--al-conc <path>
Write paired-end reads that align concordantly at least once to file(s) at <path>. These reads correspond to the SAM records with the FLAGS 0x4 bit unset and either the 0x40 or 0x80 bit set (depending on whether it’s mate #1 or #2). .1 and .2 strings are added to the filename to distinguish which file contains mate #1 and mate #2


So as output you will have 2 fastq files that contain the reads that align concordantly to your options.

For the range between pairs, take a look at --maxins option.

hope it will help

aubin

[lorenzo.C]

thanks for the help good sir, if I understand correctly --al-conc <path> will produce a new couple of fastq that fulfil the --maxins option. the maxins command will take just reads that are spaced to a maximum number of base pairs in between. the problem I have is that the second read should have the opposite orientation of the first read to be selected by maxins (if I understand correctly), is there a way to not have this option selected? in other words I need to isolate reads that are within a certain range of the other direction, no matter the orientation (both should be considered valid).