using custom scheme no sequence type is assigned

[stefcardinale]

Hello, I have built a custom scheme based on a list of fasta files of loci using the 'build_db' function, I have also given the profile with Sequence Types and their allelic profiles using the -p option (file is a .txt file).
However, when I use the scheme for calling alleles on SRR Illumina reads, the process completes fine but no Sequence Type from the profile is assigned. How can I use the scheme profile information? is it lost somewhere? I noticed that the first function creates a xx.db.profile file with the appropriate information, but apparently it is not used? Thanks for the help!

[dfornika]

Hi @stefcardinale thanks for using MentaLiST and reporting this issue. Which version of MentaLiST are you using? Could you post the commands that you used to build your db and call alleles?

[stefcardinale]

Hello, thanks for writing. I have installed Mentalist last week through manual installation following the instructions (installing Julia dependencies first).
I have created a custom scheme (C.jejuni) from a list of fasta files with the command:
mentalist build_db --db ./newScheme/newJejuni_scheme.db -k 25 -p ./jejuni_bigsdb.txt -f ./download/*.fas

This created the following files:
newJejuni_scheme.db
newJejuni_scheme.db.profile

Then I used the command to call alleles:
mentalist call -o test --db ./newScheme/newJejuni_scheme.db -1 SRR2985019_sampled.R1.fastq

The process completes fine, but I always get ST=0, is it normal? I would expect the tool identifies the closest sequence type from the scheme profiles. Thanks!

[dfornika]

Could you post your jejuni_bigsdb.txt file? Just a small portion of it should be ok, I'd just like to see the format.

[pedrofeijao]

Hi @stefcardinale , Mentalist will only assign an ST if it matches exactly the profile, it will not try to find the closest ST otherwise. Are you getting ST=0 for a genotype that matches the profile?

[slvrshot]

Could you post your jejuni_bigsdb.txt file? Just a small portion of it should be ok, I'd just like to see the format.

Is there no way to make a call based on the closest allele match at some threshold?