Working With Fastq Files In Gmap

Question

[Assa Yeroslaviz]

Hi,

I am trying to work with the gmap/gsnap tools. I have fastq files and wanted to ask whether it make sense at all to try and convert the reads into fastA format so that I can work with the gmap tool.

I think, gmap work only with fastA but not with fastq, while gsnap can work with both.

I converted the fastq to fastA which looks like that:

>HWI-ST863:138:D0WT7ACXX:5:1101:1179:1941 1:N:0:GCCAAT
ACTTACAGCATCCTCTCATCTGGGTTAACTTACTCTTGGAGGAGATGAAGGTCACTGTGCCTGGCTATCAAACCTTCCATGCTGCTCCACAGAGACAAAGA
>HWI-ST863:138:D0WT7ACXX:5:1101:1577:1894 1:N:0:GCCAAT
NTGTGAGTATGTGTGTTGTGAGAGAGAGTATGTTTGTGTGTGTGAATGTGTGTGTCTCTGTGAGTGTGTGTGTGTGAGAGTATGTTTGTGTGTGTGTGAAT
>HWI-ST863:138:D0WT7ACXX:5:1101:1718:1912 1:N:0:GCCAAT
CTGCCTCTTCACATCTCAAAAGGTGATTAACATCCGGAGATTTACATCATCTGCCAAGTCCTGCTGATCAAACAGCAGGGTAAGTATGGAGAAGGGCCACA
...

I created the indexed files with my fastA DNA file. It worked good, but when I am running the gmap, it gives me always the error message "no paths found"

and the files resulted is .nomapping .fa, which looks like that:

>HWI-ST863:138:D0WT7ACXX:5:1101:1403:1981 1:N:0:AGTCAA
 Paths (0):

 >HWI-ST863:138:D0WT7ACXX:5:1101:1470:1997 1:N:0:AGTCAA
  Paths (0):

 >HWI-ST863:138:D0WT7ACXX:5:1101:1677:1894 1:N:0:AGTCAA
 Paths (0):
 ...

Is there an explanation for this behavior? Is it possible to use this fastq-converted fastA files to run gmap?

Thanks for the help, Assa

[Sean Davis]

Just out of curiosity, why do you want to avoid using gsnap?

[Sean Davis]

Could you supply the gmap command line?

[Assa Yeroslaviz]

I am searching for large indels in the genome of my mutated mice. I wanted to try and map them to the genome to see if I can detect any large insertions/deletion with gmap or gsnap.

I don't want to avoid gsnap, I just wanted t see the differences between the mapping of the two software. I think, if it is possible to map the fatsq file to the genome, it will be easier to observe the deletions.

my gmap command is:

gmap -D gmap_index/MT_mouse -d MT_mouse --split-output= G -j 64 --fails-as-input G_R2.fa

any ideas?

[pengchy]

what about using another mapping tool, such as blat.