PE SOLiD reads alignment by bwa

[m_elena_bioinfo]

Dear users,
I have PE reads from SOLiD to align to human genome.
I have these files:

- solid_data_F3.csfasta
- solid_data_F3_QV.qual
- solid_data_F5-P2.csfasta
- solid_data_F5-P2_QV.qual

I want to convert in fastq these files by using bwa0.5.7/solid2fastq.pl
This script runs only for F3 but with F5-P2 the program doesn't run. (it says Fail to open solid_data_F5-P2_F3.csfasta)

So, if I use:
> solid2fastq.pl solid_data_ solid_data_total
I generate only one file fastq for F3 and F5-P2. It includes all the paired-end?

This fastq is in colorspace but the colors are represented as ACTG.
So to index the genome and to perform bwa alignment, have I to use -c option?

Thanks a lot,
ME

[nilshomer]

Originally Posted by m_elena_bioinfo

Dear users,
I have PE reads from SOLiD to align to human genome.
I have these files:

- solid_data_F3.csfasta
- solid_data_F3_QV.qual
- solid_data_F5-P2.csfasta
- solid_data_F5-P2_QV.qual

I want to convert in fastq these files by using bwa0.5.7/solid2fastq.pl
This script runs only for F3 but with F5-P2 the program doesn't run. (it says Fail to open solid_data_F5-P2_F3.csfasta)

So, if I use:
> solid2fastq.pl solid_data_ solid_data_total
I generate only one file fastq for F3 and F5-P2. It includes all the paired-end?

This fastq is in colorspace but the colors are represented as ACTG.
So to index the genome and to perform bwa alignment, have I to use -c option?

Thanks a lot,
ME

It looks like the script doesn't support the paired end protocol. Bug the BWA mailing list (bio-bwa-help@lists.sourceforge.net) or the author (username:lh3).

[drio]

If you want to use the script with the PE data make this change in the script:

98 #if (/^>(\d+)_(\d+)_(\d+)_[FR]3/) {
99 if (/^>(\d+)_(\d+)_(\d+)_[F3|R3|F5-P2]/) {

And also rename the F5-P2 to R3:

solid_data_F5-P2.csfasta -> solid_data_R3.csfasta
solid_data_F5-P2_QV.qual -> solid_data_R3_QV.qual

Also, bfast has a solid2fastq (in the git repo) that supports now bwa output and
handles PE data. You can use that too.

[m_elena_bioinfo]

Thanx very much for your help Drio!
I'll try and let you know if the program run!

[SoftGenetics]

Originally Posted by m_elena_bioinfo

Dear users,
I have PE reads from SOLiD to align to human genome.
I have these files:

- solid_data_F3.csfasta
- solid_data_F3_QV.qual
- solid_data_F5-P2.csfasta
- solid_data_F5-P2_QV.qual

I want to convert in fastq these files by using bwa0.5.7/solid2fastq.pl
This script runs only for F3 but with F5-P2 the program doesn't run. (it says Fail to open solid_data_F5-P2_F3.csfasta)

So, if I use:
> solid2fastq.pl solid_data_ solid_data_total
I generate only one file fastq for F3 and F5-P2. It includes all the paired-end?

This fastq is in colorspace but the colors are represented as ACTG.
So to index the genome and to perform bwa alignment, have I to use -c option?

Thanks a lot,
ME

You will loose a lot of information by converting the color space files to fasta, you would be better off aligning the solid reads to a color space reference

John

[drio]

There is information lost because of the dinucleotide 'color' encoding but the alignments are performed in CS (http://seqanswers.com/forums/showthread.php?t=5245). BWA will do a good job aligning those reads.

[SoftGenetics]

Originally Posted by drio

There is information lost because of the dinucleotide 'color' encoding but the alignments are performed in CS (http://seqanswers.com/forums/showthread.php?t=5245). BWA will do a good job aligning those reads.

We utilize a modified BWA in our NextGENe software which adds a couple of additional steps to the BWA alignment, creating a much more robust alignment, addtionally, we utilize a fully annotated color space reference so no information is lost, if you would like to try, we can supply a trial.
John

[drio]

Cool, any plans to integrate that into the main bwa repo?

[Agent47]

Thanks! Elena and drio

This was useful. i am trying to run the solid pe barcoded analysis.
I have submitted it to run just now.
I hope this works.