translating a bam file to protein sequences

Question

[Assa Yeroslaviz]

we would like to translate the genomic sequences in a bam file to its protein sequence (all six frames).

We would like to do it while keeping the structure of the bam file ( or at least a sam file), if possible.

I was wondering if anyone know about a tool to do such a task.

thanks

[JC]

That is something really uncommon because you have millions of short reads (nucleotides) in the BAM file.

If you a looking for a particular protein, it is better to do some assembly and look for your protein in the contigs.

[Assa Yeroslaviz]

Yes, I know this is uncommon und I couldn't find any solution for that yet. But I still look for one, if possible. It must not be a huge bam file, it can be one where we have extracted reads based on a bed file for specific positions. It is mainly meant to be used in identifying SAP (single AA polymorphism), so we might split a possible bam file for only the positions we need.

[GenoMax]

we would like to translate the genomic sequences in a bam file

Sequence present in each line/record?

[Assa Yeroslaviz]

Yes, We might not need it for the whole bam file (see above), but I would like to know if there is an option.

[GenoMax]

You can convert the reads into fasta format by using BBMap suite.

reformat.sh in=your.bam out=file.fa
reformat.sh in=your.bam out1=R1.fa out2=R2.fa

Then use sixpack or transeq from EMBOSS to get the translations in all frames.

[Assa Yeroslaviz]

Thanks I'll try that. I was hoping for a tool which can keep the bam/sam structure, but I'll keep looking.