Profile file for cgMLST

[lskatz]

Hi, related to #11, could you show me what a profile file should look like for cg/wgMLST? I downloaded the wgMLST scheme from the ChewBBACA website and I want to make a STing database out of it. I basically ran this inside the folder of locus fasta files:

ls *.fasta| head | \
  perl -MFile::Basename=basename -lane '
    BEGIN{print "[loci]";} 
    $n=basename($F[0], ".fasta"); 
    $n=~s/INNUENDO_cgMLST-//; 
    print join("\t", $n, $F[0]); 
    END{print "[profile]"; print "profile\tprofile.txt";}
  ' > config.txt

touch profile.txt

And then I get this error

[gzu2@monolith3 Salmonella_enterica.stringMLST]$ indexer -c config.txt
Loading sequences from sequences files:

N       Loci    #Seqs.  File
1       00031717        15      ./INNUENDO_cgMLST-00031717.fasta
2       00031718        35      ./INNUENDO_cgMLST-00031718.fasta
3       00031719        14      ./INNUENDO_cgMLST-00031719.fasta
4       00031720        42      ./INNUENDO_cgMLST-00031720.fasta
5       00031721        5       ./INNUENDO_cgMLST-00031721.fasta
6       00031722        30      ./INNUENDO_cgMLST-00031722.fasta
7       00031723        17      ./INNUENDO_cgMLST-00031723.fasta
8       00031724        17      ./INNUENDO_cgMLST-00031724.fasta
9       00031725        23      ./INNUENDO_cgMLST-00031725.fasta
10      00031726        11      ./INNUENDO_cgMLST-00031726.fasta

Total sequences loaded: 209

Loading the profiles file...
ERROR: At least 11 columns (a ST column + # loci in config file) are required in the profiles file but only 0 were found.

[ar0ch]

Hey Lee, I can pull out an example this evening but to unblock you asap the profile file needs to minimally have two lines that looks like:

ST loci1 loci2 ... lociN
1 1 1 ... 1

Which is just a single dummy ST with allele 1 for all loci.

[lskatz]

Ok I bashed it out a bit with those instructions. Thanks for that clarity. Here are the steps I took after the perl -lane step.

numLoci=$(\ls -f1 *.fasta | wc -l)
(
  echo -ne "ST\t"; 
  grep -B 999999 'profile' config.txt | grep -v profile | tail -n +2 | cut -f 1 | tr '\n' '\t'; 
  echo;
) > profile.txt
(for i in `seq 1 $numLoci`; do echo -ne "1\t"; done; echo) >> profile.txt
indexer -c config.txt

It seemed to be working but I ran into a segmentation fault. What do you think?

...
8548    00040264        49      ./INNUENDO_cgMLST-00040264.fasta
8549    00040265        30      ./INNUENDO_cgMLST-00040265.fasta
8550    00040266        144     ./INNUENDO_cgMLST-00040266.fasta
8551    00040267        21      ./INNUENDO_cgMLST-00040267.fasta
8552    00040268        35      ./INNUENDO_cgMLST-00040268.fasta
8553    00040269        12      ./INNUENDO_cgMLST-00040269.fasta
8554    00040270        33      ./INNUENDO_cgMLST-00040270.fasta
8555    00040271        105     ./INNUENDO_cgMLST-00040271.fasta
8556    00040272        71      ./INNUENDO_cgMLST-00040272.fasta
8557    00040273        38      ./INNUENDO_cgMLST-00040273.fasta
8558    00040274        50      ./INNUENDO_cgMLST-00040274.fasta

Total sequences loaded: 2898302

Segmentation fault

[lskatz]

I ran time to see if memory or something might be an issue and I guess it is saying it didn't take too much. Just about 7G.

\time index -c config.txt

24.62user 6.63system 0:38.31elapsed 81%CPU (0avgtext+0avgdata 7153812maxresident)k
0inputs+0outputs (0major+1767315minor)pagefaults 0swaps