bacpop / PopPUNK

PopPUNK 👨‍🎤 (POPulation Partitioning Using Nucleotide Kmers)
https://www.bacpop.org/poppunk
Apache License 2.0
87 stars 17 forks source link

--write-references option when running PopPUNK assign #265

Closed avrilcoghlan closed 7 months ago

avrilcoghlan commented 1 year ago

Hello,

I am using poppunk 2.6.0 on the Sanger farm.

I built a database for Vibrio cholerae using poppunk and now tried assigning some other sequences, using the command:

poppunk_assign --db chun_poppunk_db1 --model-dir chun_poppunk_db_fitted1 --query vibriowatch_fastas/input_file --output test_assign_vibriowatch_poppunk_clusters --serial --write-references

Some of the new sequences that I assigned were in already used to make the initial database, which is why I used the '--write-references' option. I found that in the output csv file of poppunk clusters, some of these were identified with '_query' added to the sequence name. However, there were other sequences that I knew were used to make the initial database, but they didn't have '_query' added to the sequence name in the output csv file, even though I would have expected them to be. I'm wondering is that a bug?

Regards, Avril

johnlees commented 1 year ago

So to summarise, you've got some queries in the initial database, not identified as queries in the final output?

Can you check in your original fit chun_poppunk_db_fitted1 that:

johnlees commented 7 months ago

Closing due to lack of response