Please help aligning SOLID PE data

[christophpale]

Hi
I've been having problem with SOLID pe data and i was hoping you can help. I am use to hiseq data but the tools for SOLID are unclear to me.
I ran my SOLID 50 and 30 bp paired reads with novoalign--that failed; then running with bwa failed because of the way the reads are listed as R5 versus R3 (I forget the exact detail).
I then used BFAST which took forever (2 weeks). then I called mutations with GATK but I only get 5000K variants when I expected over 30K (for agilent sureselect 50M).
Although it is possible the data I have is insufficiently covered, it would be good to know that I have done all the steps properly.

Could you please share your the code that you use to align the csfasta files?
for example, the reads I have have the following naming
abc_F3.csfasta
abc_R5-BC.csfasta
abc_F3_QV.qual
abc_R5-BC_QV.qual

thanks !

[gringer]

The solid reads are aligned in RF orientation (R5-BC for the reverse strand, F3 for the forward strand), so you may need to explicitly indicate this for your alignment programs. I've had trouble aligning with Bowtie paired-end for some reason, but the unpaired reads seem to align fine.

[kevleb]

i use bowtie to align paired-ends reads from SOLiD :
bowtie -f --fr -C -S -I 0 -X 500 --Q1 f3.qual --Q2 f5.qual -1 f3.csfasta -2 f5.csfasta bowtie_index_c algn.sam
Bowtie_index_c must be in color space (bowtie-build -C seq.fa seq_c)
kevin.