rename reads in fastq files

Question

[slimane.khayi]

Hi all, I have fastQ file and I need to rename it using sed command. below the explanation :

The read names in my files are

@HWI-ST365:251:D0RP0ACXX:5:1101:4471:2213#12_1
@HWI-ST365:251:D0RP0ACXX:5:1101:4471:2213#12_2

And i want to transform them in the format:

@HWI-ST365:251:D0RP0ACXX:5:1101:4471:2213#12/1
@HWI-ST365:251:D0RP0ACXX:5:1101:4471:2213#12/2

Thank you very much in advance for your help.

[iraun]

What have you tried so far?

[slimane.khayi]

based of some forum, I tried sed -i 's/_///g' myfile but I'm not pro of linux I don't know how to do..

[WouterDeCoster]

Avoid sed -i when you are not sure that your command will be the right one. If you do something incorrectly, it will corrupt your original file. When trying different sed commands, you may want to run

sed 's/from/to/g' <input> | head (to only look at the first lines)

or

sed 's/from/to/g' <input> | head | less -S ( in the case of long lines)

Answer 1

[iraun]

Well, you're very close to the solution. You only need to scape to '/' character: sed -i 's/_/\//g' should work.

Just a little advice, try to call to sed command in the following way:

cat file.fq | sed 's/_/\//g' > reformat.fq

In this way you can go back to the original input file in the case that something has gone wrong. In my opinion it is a good practice.

[venu]

Just as an alternative tr '_' '/' < file.fq > new_file.fq

[slimane.khayi]

Thank you I'll try this.

Answer 2

[michael.ante]

Check your fastq format. If you have Phred +64 (Illumina 1.3 or 1.5) you can run into a encoding problem: in Phred +64, '_' is a valid encoding for a quality score, '/' is not. Thus, you'll need to check if you are in the header-line or not (e.g. using awk: awk '{if(NR%4==1){gsub(/_/,"/")}; print}'' )

[untitpoi]

I agree that you should verify the absence of _ in your quality sequence before to simply go for a sed 's/_/\//g'. because if there is you will change all your quality score coded _ by a new existing score \