Paired-end reads into two different files

[Fad2012]

Hello all

I am using a software called MetaSim to generate reads based on given genome sequences. My problem is that the reads i produced are paired-ends in one multifasta file. the paired reads look like the following:

>r1.1 |SOURCES={GI=61252,fw,624-696}|ERRORS={67:A}|SOURCE_1="Human poliovirus 1 Mahoney" (8915ea1a18cb58f4a76d99a56ece2a9018e105bc)
ATTGGCCATCCGGTGAAAGTGAGACTCATTATCTATCTGTTTGCTGGATCCGCTCCATTGAGTGT
GTATACT
>r1.2 |SOURCES={GI=61252,bw,789-861}|ERRORS={}|SOURCE_1="Human poliovirus 1 Mahoney" (8915ea1a18cb58f4a76d99a56ece2a9018e105bc)
GCGTTACTAGCTGAATCTCTATAATAATTAATGGTGGTGTAATTAATGGTAGAACCACCATACGC
TCTATTT
>r2.1 |SOURCES={GI=61252,fw,6323-6395}|ERRORS={}|SOURCE_1="Human poliovirus 1 Mahoney" (8915ea1a18cb58f4a76d99a56ece2a9018e105bc)
CCACCAGTGCTGGCTACCCTTATGTAGCAATGGGAAAGAAGAAGAGAGACATCTTGAACAAACAA
ACCAGAG
>r2.2 |SOURCES={GI=61252,bw,6502-6574}|ERRORS={}|SOURCE_1="Human poliovirus 1 Mahoney" (8915ea1a18cb58f4a76d99a56ece2a9018e105bc)
AGCATATAGGTTCCCAAAAGCCATTCTCATTGCCACTGAGTCATTCAAACTAGAAGCTTCAATTA
ATCTGGA

I want the backward reads in a separated file, and the forward reads in another file. Does anyone have any idea about a perl script or any other tool to separate these two reads into two different files; then i will be able to align them using Bowtie or any other alignment tool.

Many thanks

[dpryan]

You could just use grep. Something like this should work:

grep -A 1 ".1 " reads.fa > reads_1.fa
grep -A 1 ".2 " reads.fa > reads_2.fa

[vivek_]

Originally Posted by dpryan

You could just use grep. Something like this should work:

grep -A 1 ".1 " reads.fa > reads_1.fa
grep -A 1 ".2 " reads.fa > reads_2.fa

This should print only the headers and not the sequence right?

[dpryan]

Originally Posted by vivek_

This should print only the headers and not the sequence right?

No, check the man page for grep.

[vivek_]

My bad! Did not notice the -A 1 at the beginning.

[Fad2012]

Hello

I still get both 1.1 and 1.2 in the output file. Does it have to do with the extension of my file which is .fna, this how i am doing it; in the terminal i am in the directory where my file is, my file is: test.fna:

grep -A 1 ".1 " test.fna > reads_1.fna
grep -A 1 ".2 " test.fna > reads_2.fna

Many thanks for your help

[dpryan]

Originally Posted by Fad2012

Hello

I still get both 1.1 and 1.2 in the output file. Does it have to do with the extension of my file which is .fna, this how i am doing it; in the terminal i am in the directory where my file is, my file is: test.fna:

grep -A 1 ".1 " test.fna > reads_1.fna
grep -A 1 ".2 " test.fna > reads_2.fna

Many thanks for your help

Ah, I see that it's treating "." as a match to any character, I should have foreseen that. Try instead:

grep -A 1 "\.1 " test.fna > reads_1.fna
grep -A 1 "\.2 " test.fna > reads_2.fna

I should note that, at least under Ubuntu, this results in each read being separated by a "--" line. If that happens for you too, then just add the "--no-group-separator" option (i.e, grep -A 1 --no-group-separator "\.1 " test.fna > reads_1.fna).

[Fad2012]

Thanks very Much dpryan, it is working now. But yes i do have the separator and when i try to use it, the terminal says: unrecognized option `--no-group-separator'. I am not sure what the reason could be! My terminal is Version 2.3 (309).

Many thanks!

[GenoMax]

You can add a second grep to remove the "--"

grep -v "\--" reads_1.fna > reads_1_final.fna

or you could just add the second grep to the original command

grep -A 1 "\.1 " test.fna | grep -v "\--" > reads_1.fna

[Fad2012]

Thanks Very much, it is perfectly working!

[Smaugt]

Hi all,

i have the same problem reported here but if do "grep -v "\--" > reads_1.fna" I also remove lines starting with "--"...
Thanks in advance

[GenoMax]

Originally Posted by Smaugt

Hi all,

i have the same problem reported here but if do "grep -v "\--" > reads_1.fna" I also remove lines starting with "--"...
Thanks in advance

Can you post a small example from your file? Why do you have lines that start with "--"?

[Smaugt]

Thanks for ur soon response...

OK, I explain better.... I have something like this

@IC5OSZZ01DTNH9/1
GTTGTCGTGGCTCATGTTCGAGTTATCCATTTGTGCGAATGCGCCTGCTGATACCATG
+IC5OSZZ01DTNH9/1
--GIIIIIIIIIIII444IIIIEEIEIII444IIIIIIIIIIEGGHIIIII@888EII
--

if I remove the -- with "grep -v "\--" > reads_1.fna" it'll remove the qualities row too...
I hope you can help me..

Thanks!

[GenoMax]

What OS are you using?

Did you try the "--no-group-separator" option mentioned by Devon in post #7 instead of the second grep?

[dpryan]

Originally Posted by Smaugt

Thanks for ur soon response...

OK, I explain better.... I have something like this

@IC5OSZZ01DTNH9/1
GTTGTCGTGGCTCATGTTCGAGTTATCCATTTGTGCGAATGCGCCTGCTGATACCATG
+IC5OSZZ01DTNH9/1
--GIIIIIIIIIIII444IIIIEEIEIII444IIIIIIIIIIEGGHIIIII@888EII
--

if I remove the -- with "grep -v "\--" > reads_1.fna" it'll remove the qualities row too...
I hope you can help me..

Thanks!

Keep in mind that the grep commands that GenoMax and I had mentioned were for use with fasta files, where this situation wouldn't occur. As GenoMax suggested, try the "--no-group-separator" option. If that doesn't work, you can probably still use grep, you just need to change

grep -v "\--"

into something like

grep -v "^--$"

or something like that. The idea is to instead do the inverse match for whole lines that are exactly "--".