het. and hom. SNP？

[Colorful_Seq]

HI,
I use MAQ software mapping the reads.
Is there anybody know how to get the het and hom SNPs from MAQ software?

Thx!

[qtrinh]

Hi,
Running the command 'maq cns2snp in.cns > out.snp' should give you the info you are looking for - see http://maq.sourceforge.net/maq-manpage.shtml for output column descriptions.

Q

[Colorful_Seq]

Thx Q...
I already Runing the command 'maq cns2snp in.cns > out.snp'.
Now I write a simple perl script to split the snp file to hom snp and het snp.

[qtrinh]

No ... no ... no perl ... a one-liner bash with awk will do. Let me know if you need the one-liner ... :-)

Q

[Colorful_Seq]

Really?
I am sorry,I am not familiar with Linux command.
But I'd like to know!
Thanks a million!

[qtrinh]

cat MAQ_SNP_FILE | awk '{ if (($4!="A") && ($4!="C") && ($4!="G") && ($4!="T") ) { print $0 } }' > het.snps



cat MAQ_SNP_FILE | awk '{ if (($4=="A") || ($4=="C") || ($4=="G") || ($4=="T") ) { print $0 } }' > homo.snps


$4 means column number 4 in your MAQ_SNP_FILE.

Q

[Colorful_Seq]

Ha~cool!
It's works...
It's better than perl!
faster and just a liner~

Thanks a million!
Thank you Q...

[westerman]

Perl is also good

Of course perl can be a one-liner as well. And, IMHO, a cleaner one than awk.

perl -nle '@z = split /\t/; print if $z[3] !~ /[ACGT]/' MAQ_SNP_FILE > het.snps

perl -nle '@z = split /\t/; print if $z[3] =~ /[ACGT]/' MAQ_SNP_FILE > homo.snps


Basically, the 'perl -nle' means:

'n' ==> loop through the file
'l' ==> remove and add newlines as needed
'e' ==> execute the following script.

The script is in two parts:

@z = split /\t/ ==> takes the line and splits it along tabs into an array called '@z'
print if the 4th element of the array (called [3] in the perl 0-based array structure) is not (hets) or is (homo) one of A, C, G, or T.

All-in-all the perl method seems more clean to me since the character to look for can be bundled into the square brackets; e.g., [ACGT]. But then I am mainly a Perl programmer.

[qtrinh]

Yes, I agree with westerman, Perl is also good. I found myself using a bit of Perl, awk, and sed everyday.

Q