k-mer error while running strainscan

[AK-25FL]

Hi,

Thank you for developing this pipeline. This is very helpful for strain-level identification.
I used it for strain-level identification within the species. It worked very well. Then, I tried to identify strains from three highly related species. After building the database while running the strainscan pipeline it gave the following error. I don't know if this has something to do with the update or any other error related to k-mer.
Will appreciate your help.

Thanks,

Traceback (most recent call last):
File "/blue/env/bin/strainscan", line 10, in
sys.exit(main())
File "/blue/env/lib/python3.7/site-packages/StrainScan/StrainScan.py", line 198, in main
cls_dict=identify.identify_cluster(in_fq,db_dir+'/Tree_database',[0.1,0.4,1])
File "/blue/env/lib/python3.7/site-packages/StrainScan/library/identify.py", line 406, in identify_cluster
match_results = jellyfish_count(fq_path, db_dir)
File "/blue/env/lib/python3.7/site-packages/StrainScan/library/identify.py", line 98, in jellyfish_count
match_results[kmer_index_dict[temp[0]]] = int(temp[1])
KeyError: 'TGCGCACCTTTGCGCTGCTGGTGGCGGTGCA'

[liaoherui]

Hi! Thank you for using StrainScan.

Could you please share the input genome files causing the problem with us? With the genomes, we will investigate the issue immediately and provide you with both the potential cause and the solution. Thanks!

[AK-25FL]

Hi,

Thanks for replying. Here's the link for the genome files.

https://drive.google.com/drive/folders/12A3PGOZ-FhLiAnWhdQ7h9j9FuWC9kZ2Z?usp=sharing

Please let me know if you need additional information.

Thanks,

[liaoherui]

Sorry for the late response—I’ve been quite busy recently.

I changed the names of your input genomes before running the build scripts, and I was able to build the database successfully. You can find the pre-built database for your provided genomes here:

https://drive.google.com/file/d/14LWKDAsPUvDzXqsC30_V_-3MaZwIY3Yn/view?usp=sharing

If you have any other questions or issues, please let me know.

[AK-25FL]

Hi,

Thanks, I appreciate your time.

I used the db you provided to identify known strains (those already present in the db), but I encountered the same error as last time. Below is the command I used and the error I got. I have also attached the reads that I am trying to identify in case there is an issue with the input. However, I don't think that's the case since I was able to detect them when I built db constituting strains from only one species.

Command: "strainscan -i R1.fq -j R2.fq -d db -o out"

Error:

Traceback (most recent call last):
File "/blue/env/bin/strainscan", line 10, in
sys.exit(main())
File "/blue/env/lib/python3.7/site-packages/StrainScan/StrainScan.py", line 198, in main
cls_dict=identify.identify_cluster(in_fq,db_dir+'/Tree_database',[0.1,0.4,1])
File "/blue/env/lib/python3.7/site-packages/StrainScan/library/identify.py", line 406, in identify_cluster
match_results = jellyfish_count(fq_path, db_dir)
File "/blue/env/lib/python3.7/site-packages/StrainScan/library/identify.py", line 98, in jellyfish_count
match_results[kmer_index_dict[temp[0]]] = int(temp[1])
KeyError: 'CCCCGGGCGCCTTTGCCGACACCCTGCATCT'

Thanks,
Will appreciate your help.

[liaoherui]

Oh, I see... this appears to be an issue with the identification process rather than database construction. Could you please provide both R1.fq and R2.fq files in your command line? Thanks a lot!

[AK-25FL]

Here's the link for R1.fq and R2.fq

https://drive.google.com/drive/folders/14Sr7H7HDEMblrQq3meRZNie7EnAiWrrS?usp=drive_link

[liaoherui]

Hi, thanks for providing the test data! Could you please try the latest GitHub version of StrainScan? You can just clone the github and run the code StrainScan.py under your "strainscan" conda environment.

In our test, we find that python StrainScan.py -i tmp/R1.fq -j tmp/R2.fq -d tmp/db/ -o tmp/res/ works well (tested with the database I built before).

[AK-25FL]

Hi,

Sorry for replying late.
It's working.

Thanks for your help.