mag: add pipeline compatible mini test database for upcoming CAT_pack #1408
Conversation
| ## Generate database using CAT_pack prepare | ||
| CAT_pack prepare --db_fasta input_files/sequence_fixedheaders.txt --names input_files/names_reduced.dmp --nodes input_files/nodes_reduced.dmp --acc2tax input_files/accession2taxid_reduced.dmp --db_dir test/ | ||
|
|
||
| ## Test using uncompressed contigs from metaspades assembly (note: --no_stars was required for some reason but seesms to only occur when we have a single genome in there possible..) |
There was a problem hiding this comment.
Will this be a problem using it when testing in #mag? Or will it be OK to just add --no-stars to the test config?
There was a problem hiding this comment.
The latter as far as I can tell :)
The flag was added for this purpose according to GitHub
There was a problem hiding this comment.
afaik Maxbin2 will force a split into the contigs, it just assumes there are 2+ genomes (at least that happened a few years ago). But thats fine I think.
I agree but would need @HadrienG @skrakau @d4straub to do some data archaeology for us 😅 Although I was speaking with @muabnezor about the long read only stuff he adding to mag and we were wondering if it would be worth replacing the test data now to something less mysterious and has paired short and long read data before we implement nf-test... |
There was a problem hiding this comment.
I agree that we should document where the data comes from. I'm going to speak with Sabrina about that.
Regarding whether that is related to nf-test: I dont think that matters whatsoever, because no normal user is running the tests with nf-test on old versions (usually -profile test without nf-test), and devs might be fine that a past release fails with nf-test because they are usually just interested in the dev branch.
| curl "https://www.ncbi.nlm.nih.gov/sviewer/viewer.cgi?tool=portal&save=file&log$=seqview&db=nuccore&report=fasta_cds_aa&id=1992822979&extrafeat=null&conwithfeat=on&hide-cdd=on&ncbi_phid=CE8C15326D6BB8C10000000006490560" -o sequence.txt | ||
|
|
||
| ## use my scripts to filter NCBI nodes/names/acc2taxid files to just a given taxid (basically fancy iterative greps) | ||
| bash ~/bin/taxdmp_filter.sh 817 ## for nodes/naames |
There was a problem hiding this comment.
| bash ~/bin/taxdmp_filter.sh 817 ## for nodes/naames | |
| bash ~/bin/taxdmp_filter.sh 817 ## for nodes/names |
| sed 's/lcl|//g;s/_/ /2' sequence.txt > sequence_fixedheaders.txt | ||
|
|
||
| ## Generate database using CAT_pack prepare | ||
| CAT_pack prepare --db_fasta input_files/sequence_fixedheaders.txt --names input_files/names_reduced.dmp --nodes input_files/nodes_reduced.dmp --acc2tax input_files/accession2taxid_reduced.dmp --db_dir test/ |
There was a problem hiding this comment.
| CAT_pack prepare --db_fasta input_files/sequence_fixedheaders.txt --names input_files/names_reduced.dmp --nodes input_files/nodes_reduced.dmp --acc2tax input_files/accession2taxid_reduced.dmp --db_dir test/ | |
| CAT_pack prepare --db_fasta input_files/sequence_fixedheaders.txt --names input_files/names_reduced.dmp --nodes input_files/nodes_reduced.dmp --acc2tax input_files/accession2taxid_reduced.dmp --db_dir test/ |
| ## Generate database using CAT_pack prepare | ||
| CAT_pack prepare --db_fasta input_files/sequence_fixedheaders.txt --names input_files/names_reduced.dmp --nodes input_files/nodes_reduced.dmp --acc2tax input_files/accession2taxid_reduced.dmp --db_dir test/ | ||
|
|
||
| ## Test using uncompressed contigs from metaspades assembly (note: --no_stars was required for some reason but seesms to only occur when we have a single genome in there possible..) |
There was a problem hiding this comment.
afaik Maxbin2 will force a split into the contigs, it just assumes there are 2+ genomes (at least that happened a few years ago). But thats fine I think.
|
About the origin of the test data: It seems to be from @HadrienG and there might be a discussion in slack about it. |
No description provided.