SHRiMP - how to obtain unique mapped reads?

[idonaldson]

Hi,

Can any SHRiMP users tell me how they obtain unique mapped reads (those mapped reads that only map to one position in the genome). I am mapping 50bp reads to the human genome. I have run rmapper-cs using the 'fast' and '50bp' settings.

Thanks!

Ian

[idonaldson]

nudge

nudge for second viewing, thanks!

[nilshomer]

nudge for second viewing, thanks!

You will need to write your own post-alignment filtering script. You can bug the authors to get the output to SAM format, which would make things easier.

[sci_guy]

SHRiMP now can output in SAM

You will need to write your own post-alignment filtering script. You can bug the authors to get the output to SAM format, which would make things easier.

The latest 6 Oct (V1.30) version has a shrimp2sam.pl script in the /Utils folder.

[Torst]

Hi,
Can any SHRiMP users tell me how they obtain unique mapped reads (those mapped reads that only map to one position in the genome). I am mapping 50bp reads to the human genome. I have run rmapper-cs using the 'fast' and '50bp' settings.

This perl code will take 'hits.txt' on stdin and write 'unique_hits.txt' on stdout.

#!/usr/bin/perl -w
my %seen;
while (my $line = <>) {
  my @f = split m/\t/, $line;
  next unless @f == 10;
  next if $seen{ $f[0] }++;
  print $line;
}

[idonaldson]

Thanks for your input!

[sci_guy]

You will need to write your own post-alignment filtering script. You can bug the authors to get the output to SAM format, which would make things easier.

@Nils; I am amused The SAM munging Perl script in the 1.3.0 SHRiMP distribution is actually one authored by you.

[nilshomer]

@Nils; I am amused The SAM munging Perl script in the 1.3.0 SHRiMP distribution is actually one authored by you.

I wrote it to compare aligners a while ago for a paper that is being published tomorrow (yay). Needless to say, I have done a lot of aligner comparisons.

Unfortunately the SHRiMP to SAM script does not support paired end reads...

[ulisses de padua]

I do rmmaper -cs and I am using sensitive 50bp...
my dates are mate pair, and I do it for _F3 file and for _R3file... The results are in two
different folders.

I am using Shrimp 1.3.2....

How I obtain the unique file for run the probcalc?


I would like to receive a answer by my personal email too: email: [email protected]

Thanks regards

[rahul.m.dhodapkar]

I am trying to run SHRiMP2 on paired-end data. I have run the indexing step, and am using

./gmapper-ls /Test_seq/test.fa -L hg18_combined.fa -V -N 8 -p opp-in >/Out_seq/shrimptestoutput

as my command. The index loads fine, and the reads file also loads, but then SHRiMP doesn't actually run and i get 0 mappings. Does anyone know how to fix this problem?

Many thanks.

[misko]

Hi Rahul,
I believe the command you wish to run is,
./gmapper-ls -L hg18_combined.fa -V -N 8 -p opp-in /Test_seq/test.fa >/Out_seq/shrimptestoutput
If you wish to have SAM output you can use,
./gmapper-ls -E -L hg18_combined.fa -V -N 8 -p opp-in /Test_seq/test.fa >/Out_seq/shrimptestoutput
If you wish to output the unaligned reads in SAM format or into a separate file, please email me ([email protected]) and i can supply you with the unreleased version that supports this feature!

Misko