added kraken 1.1.1 to roary 3.13.0 dockerfile

[kapsakcj]

To address #254

Docker image is available temporarily on dockerhub under my personal repo kapsakcj/roary:3.13.0-plus-kraken

https://hub.docker.com/r/kapsakcj/roary/tags?page=1&ordering=last_updated

@erinyoung Can you try it out and let me know how it goes? Would appreciate a quick test (I don't have any data handy to do this), especially run via singularity.

One thing to note is that I'm trying something new. I've set LC_ALL=C.UTF-8 instead of the usual LC_ALL=C. I've read that this will prevent headaches down the road. Something to do with encoding or something....

If you have perl issues, we can revert back to LC_ALL=C before merging this PR and rebuilding the staphb/roary:3.13.0 docker image

[erinyoung]

I'm testing now. Will let you know if I run into issues.

[erinyoung]

Roary worked okay, and kraken didn't throw any errors per se, but it also didn't work.

It throws an error while trying to run kraken :

kraken-report: unable to find minikraken2_v2_8GB_201904_UPDATE in $KRAKEN_DB_PATH (undefined)

Even if I define KRAKEN_DB_PATH in my script or in the container, it still throws this error.

I am unsure as to why this is, so I'll be looking into it next week-ish.

[kapsakcj]

Are you perhaps using a kraken2 database ? If so, I don't think that will be compatible with kraken v1

[erinyoung]

I want to look into that.

I only got as far as trying kraken's 'minikraken_20141208' database with the same results. For whatever reason, the environmental variable KRAKEN_DB_PATH is never found.

I want to make sure that kraken itself works in the container - outside of roary.

[kapsakcj]

OK I was able to test out this container with 3 salmonella Hadar assemblies and got the optional kraken feature to run successfully.

It seems that roary is picky about the path to the kraken1 database and requires an absolute path to the directory containing the file database.kdb. If you use a relative path, roary will not be able to locate the database.kdb file

The following command worked. Fed in 3 .gff files into Roary
roary -p 4 -qc -k ${PWD}/minikraken_20171013_4GB -f roary-kraken-container-test-take2 -e -n -v prokka-GCA_01*/*.gff

Relevant bits of STDOUT:

2022/01/16 17:49:21 Running Kraken on each input assembly
2022/01/16 17:49:21 sed -n '/##FASTA/,//p' /data/prokka-GCA_011245895.1/GCA_011245895.1.gff | grep -v '##FASTA' > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011245895.1.fna
2022/01/16 17:49:21 sed -n '/##FASTA/,//p' /data/prokka-GCA_011348195.1/GCA_011348195.1.gff | grep -v '##FASTA' > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011348195.1.fna
2022/01/16 17:49:21 sed -n '/##FASTA/,//p' /data/prokka-GCA_015698805.1/GCA_015698805.1.gff | grep -v '##FASTA' > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_015698805.1.fna
2022/01/16 17:49:21 parallel --gnu -j 4 < /tmp/OErLwtcnI9
2022/01/16 17:49:21 kraken --fasta-input  --preload  --db /data/minikraken_20171013_4GB --output /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011245895.1.kraken /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011245895.1.fna  > /dev/null 2>&1
2022/01/16 17:49:21 kraken --fasta-input  --preload  --db /data/minikraken_20171013_4GB --output /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011348195.1.kraken /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011348195.1.fna  > /dev/null 2>&1
2022/01/16 17:49:21 kraken --fasta-input  --preload  --db /data/minikraken_20171013_4GB --output /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_015698805.1.kraken /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_015698805.1.fna  > /dev/null 2>&1
2022/01/16 17:49:21 parallel --gnu -j 4 < /tmp/o3Vul9C0G_
2022/01/16 17:49:48 kraken-report --db /data/minikraken_20171013_4GB /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011245895.1.kraken > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011245895.1.kraken.report
2022/01/16 17:49:48 kraken-report --db /data/minikraken_20171013_4GB /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011348195.1.kraken > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_011348195.1.kraken.report
2022/01/16 17:49:48 kraken-report --db /data/minikraken_20171013_4GB /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_015698805.1.kraken > /data/roary-kraken-container-test-take2_1642355359/EhYR7nu1Z9/GCA_015698805.1.kraken.report

Output kraken QC file

$ cat roary-kraken-container-test-take2_1642355359/qc_report.csv
Sample,Genus,Species
GCA_011245895.1,Salmonella,Salmonella enterica
GCA_011348195.1,Salmonella,Salmonella enterica
GCA_015698805.1,Salmonella,Salmonella enterica

The above command produced all output roary files as expected

[erinyoung]

I had some time today, and I can confirm that it works with kraken databases. My kraken database had been corrupted when I tested it prior. My apologies.

First, I pulled the container with singularity

singularity pull --name roary_kraken docker://kapsakcj/roary:3.13.0-plus-kraken

Then I ran it

singularity exec --bind /Volumes/IDGenomics_NAS/Data/kraken1/minikraken_20171013_4GB:/kraken_db --bind /Volumes/IDGenomics_NAS/ARLN/wastewater/Citrobacter/gff/test:/data roary_kraken roary -p 30 -qc -k /kraken_db -f /data/roary_testing  -e -n -v /data/F16.gff  /data/F18.gff  /data/F20.gff  /data/F22.gff  /data/F36.gff  /data/F42.gff  /data/F52.gff  /data/F54.gff /data/F55.gff  /data/F59.gff

And it worked!

$ cat /Volumes/IDGenomics_NAS/ARLN/wastewater/Citrobacter/gff/test/roary_testing/qc_report.csv 
Sample,Genus,Species
F16,Citrobacter,Citrobacter freundii
F18,Citrobacter,Citrobacter braakii
F20,Citrobacter,Citrobacter farmeri
F22,Citrobacter,Citrobacter freundii
F36,Citrobacter,Citrobacter freundii
F42,Citrobacter,Citrobacter freundii
F52,Citrobacter,Citrobacter freundii
F54,Citrobacter,Citrobacter braakii
F55,Citrobacter,Citrobacter farmeri
F59,Citrobacter,Citrobacter freundii

I do have a concern, however, in that I used to use it with kraken2 databases, and I'm not sure how to add in that functionality. There doesn't seem to be any documentation about that, however.

[erinyoung]

Works for me as well

[kapsakcj]

Glad to hear it ran successfully!

I do have a concern, however, in that I used to use it with kraken2 databases, and I'm not sure how to add in that functionality. There doesn't seem to be any documentation about that, however.

Huh. I thought Roary pre-dated kraken2, so I didn't think there was support for it. Does that required kraken2 to be installed or can you use kraken1 with a kraken2 database?

I was going to merge this PR and build on dockerhub since it's running properly with kraken1, but we can wait if you'd like to try it out with a kraken2 database. Let me know what you'd prefer!

[erinyoung]

I think it's fine to merge it. I am confused as well. My notes have roary using the kraken2 database, but I can't seem to replicate those. I'm willing to move on with my life.

[kapsakcj]

I'm willing to move on with my life.

agreed 😆

I'll merge and re-build the docker image on dhub and quay. Planning on re-building the staphb/roary:3.13.0 and staphb/roary:latest images

Thanks for testing and I hope this is useful for you and others!