FASTQ header correction to remove .1 and .2 from paired end - SRA files

Question

[genomics_buddy]

From this step as raw data:

@ERR797585640.1.1 1 length=151
ATATCGGGACGTGCGATCAGCACAAGCCGGCTATCGGTGTGGCGC

I have converted till this:

Read 1:

@ERR797585640.1.1
ATATCGGGACGTGCGATCAGCACAAGCCGGCTATCGGTGT

Read 2:

@ERR797585640.1.2
ATATCGGGACGTGCGATCAGCACAAGCCGGCTATCGGTGT

I have taken ERR and SRA samples for RnD.

But I want to keep only format as:

Read 1:

@ERR797585640.1
ATATCGGGACGTGCGATCAGCACAAGCCGGCTATCGGTGT

Read 2:

@ERR797585640.1
ATATCGGGACGTGCGATCAGCACAAGCCGGCTATCGGTGT

Kindly provide any commands or suggestion to remove and make changes for fastq headers

[Pierre Lindenbaum]

I have converted till this:

how ? what was the command ?

[genomics_buddy]

I have used:

sed '/^@/ {s/ .*//}' SAMPLE to convert till @ERR797585640.1.1

But i need further guidance for next step

[Pierre Lindenbaum]

you cannot use '@' as the signal for the first line of a fastq file as @ can also be a symbol of quality

[Ram]

Please stop using "convert", you're using it wrong and your sentences don't make any sense.

Answer 1

[b.contreras.moreira]

Not sure it's a good idea to name both reads the same, but here's a perl oneliner to change your FASTQ headers:

perl -lne 'if(/^(@[^\.)]+\.\d+)\.\d/){ print "$1" } else { print }'

If you want to make sure only headers are affected, in case quality line starts with at:

perl -lne 'if($.%4==1 && /^(@[^\.)]+\.\d+)\.\d/){ print "$1" } else { print }'

[dsull]

Some tools require both reads to have the same name (or least they want the part of the read name before the first whitespace to be the same).

[noodle]

IMPORTANT SIDE NOTE: This can break SRA download tools if you upload your files with identical read names. Surprisingly NCBI allows(ed?) this....I hope they changed it by now.