sqm_mapper: reuse mapped reference genome possible?

[dutchscientist]

I am using sqm_mapper.pl to clean up samples from host sequences, in the current case a combined human and pig genome. That works great, but every time I start a new batch of samples, the sqm_mapper.pl starts re-making the reference from the fasta file, which takes quite a bit of time for a 5.4 GB haploid sequece.

I had a look at the options with sqm_mapper.pl -h, and don't see the option to use an already generated reference. Is it possible to do that? If not, a suggestion to add that to any new version, it is a really useful tool :)

[jtamames]

Hello
Rename your sqm_mapper.pl to sqm_mapper.pl.original
Then copy the attached new sqm_mapper.pl script (gunzip it first) to the sqm_mapper.pl directory (utils)
The script should now accept an option "-idx ", where path is the location of your previously created references.
Let me know if it works
Best,
J

sqm_mapper.pl.gz

[dutchscientist]

You're a star! I will give that a try, much appreciated.

[dutchscientist]

Unfortunately it does not work. Sqm_mapper still demands a -r input, and then starts making a new one, but now in the folder where it was started instead of the name/temp folder where it did it previously.

This was run first:
sqm_mapper.pl -r human_pig_genome.fna -s Trento_decontamination_part3.samples -f raw -o sqm_mapper_human_pig_trento3 --filter -t 20

And then to re-use the index:
sqm_mapper.pl -r human_pig_genome.fna -idx human_pig_genome_indexed -s Trento_decontamination_part3.samples -f raw -o sqm_mapper_human_pig_trento3_reuse --filter -t 20

With the latter setting, it still starts indexing, but now with the name of the folder but in the place where it is initiated.

Just to check, would the fna file need to be in the folder with the index files?

[jtamames]

Ok, I will check it. Quick fixes often don´t work :)
Sorry about that
Best,
J

[dutchscientist]

No worries, really appreciate your efforts!