abricate and prokka - Githubissues

tseemann / abricate

:mag_right: :pill: Mass screening of contigs for antimicrobial and virulence genes

GNU General Public License v2.0

364 stars 90 forks source link

abricate and prokka #136

Closed SumeetTiwari07 closed 4 years ago

SumeetTiwari07 commented 4 years ago

Hi, I wrote a python script which converts the abricate output to the binary gene presence absence matrix. Then later i have decided the extended the same script to extract the sequences for those genes from the respective genomes. In the process i have found that some AMR gene reported to be present by abricate in the genome is absent in the output of prokka (.gff3 ) file. I didn't understand why this happened ? is there any specific reason behind this?

tseemann commented 4 years ago

Abricate and Prokka won't give the 'same' answer because they

don't use the exact same database
don't use the same gene finding system (blastn vs prodigal/blastp)
don't use the same cutoffs (--mincov/--minid vs --evalue/--coverage)

P.S. did you consider using the abricate --summary command first?

SumeetTiwari07 commented 4 years ago

Thank you !! i understood the reason now.

Yes i tried the summary command but it will gives the identity of the gene in the columns where i wanted to have 0 and 1 indication and another file giving me the exact the number of the genes in same format as produced by abricate --summary.
Now my script is working fine.

Thank you once again