de novo assembly using unmapped reads from tophat

[ccstaats]

Hello
There is a way to use the unmapped reads from tophat 1.4.1 to build a de novo assembly?
I ve been looking for programs to do this and I found SOAPdenovo.
But first, I tried to take a look at the reads (for blast analysis) and I cant convert the unmapped.fq.z tophat output.
I tried the fq_all2std.pl script but there is no fasta output.
Any clues?
Charley

[gringer]

It is most likely the errors are because the program cannot handle a compressed fastq file. Try unzipping the file first:

gunzip unmapped.fq.z

[ccstaats]

Thank you
Works nice.
Charley

[leifive]

Hi, ccstaasts,
How is your result of de novo using unmapped reads from tophat? I combined the pair-end and single-end unmapped reads, then assembled with CLC bio, unfortunately it returned numerous contigs, I have no idea how to deal with them. In addition, did SOAPdenovo work well? I have not tried it yet.

[ccstaats]

Hi, leifive.
Sorry for the long time to answer.
In fact, I used velvetg and velveth
I've got some good results results. Found some genes that should be there. They are not, as the genome is not finished.
For that, I used the protocol described in the address:

http://ged.msu.edu/angus/tutorials-2011/mrnaseq-assembly.html

Hopefully it will work for you.
charley