seeded genome assembly

[jrenart]

I am constructing a db for sequences of a specific vertebrate protein (less than 200 aa, and about 600 bp). There are a lot of genomes already sequenced but with only the raw data available. Assembly all of them just for one protein is a waste of time and computer power (that I don't have). The program PRICE (Ruby et al 2013, GGG, 3:865) does the job (assembly starting with a seed) and works well. The problem with PRICE is that it only works with paired-end sequences. Is there other software that could be used for other platforms, starting with a seed?
Thanks for the help. Jaime

[Brian Bushnell]

Why don't you map to the gene and just assemble the reads that map? For example:

bbmap.sh in=reads.fq ref=gene.fasta outm=mapped.fq minid=0.5 k=11 nodisk

Alternately, you can probably trick PRICE by reverse-complementing single-ended reads to pretend that they are read 2. For example:

reformat.sh in=read1.fq out=read2.fq rcomp

[jrenart]

Thank you for the ideas! I'll try them and post the outcome. Jaime