Closed nagoue closed 5 years ago
Hi, Have you tried to restart Galaxy? Sometimes, it seems to be needed when you're adding a first entry into an empty data table
Hi, Yes I also tried a Galaxy restart at each of the following steps:
I also over checked that fields in loc files are tab separated. Thanks,
Ok... strange! In the admin section of the galaxy instance, you have a "Data tables" section, can you tell me what are the filenames for each of the problematic tables?
I have (too) many lines in that section:
- blastdb /xxx/tools/tool-data/blastdb.loc
- blastdb /xxx/tools/tool-data/toolshed/repos/iuc/data_manager_manual/6524e573d9c2/blastdb.loc
Those 2 blastdb.loc files contain the same info.
- blastdb_p /xxx/tools/tool-data/blastdb_p.loc
- blastdb_p /xxx/tools/tool-data/toolshed/repos/devteam/ncbi_blast_plus/e25d3acf6e68/blastdb_p.loc
Those 2 blastdb_p.loc files are empty. I re-run biomaj as I had a mistake in biomaj2galaxy command (correct path is "blastdb:blast/protein/Anopheles_gambiae_protein:Anopheles gambiae protein")
- bwa_indexes /xxx/tools/tool-data/bwa_index.loc
- bwa_indexes /xxx/tools/tool-data/toolshed/repos/iuc/data_manager_manual/6524e573d9c2/bwa_index.loc
Those 2 bwa_index.loc files contain the same info.
- bwa_mem_indexes /xxx/tools/tool-data/bwa_mem_index.loc
- bwa_mem_indexes /xxx/tools/tool-data/toolshed/repos/devteam/bwa/8d2a528a9513/bwa_mem_index.loc
Those 2 bwa_mem_index.loc files are empty.
I did cleaning in shed_tool_data_table_conf.xml file which had multiple table definitions with similar name, keeping the ones coming from data_manager_manual and restarted Galaxy. No change
I did a metadata reset, not sure it is connected but ... I noticed Nucleotide BLAST database gives access to all indexes (i.e. Anopheles gambiae, Anopheles gambiae protein) but no index is accessible from Protein BLAST database. Do I miss something in my Galaxy post-process to well-separate nucleotide and protein index?
Ah yes! I had not seen it, but in your bank properties file, it should be this way:
"blastdb_p:blast/Anopheles_gambiae_protein:Anopheles gambiae protein"
instead of:
"blastdb:blast/Anopheles_gambiae_protein:Anopheles gambiae protein"
Cool! It's working. Thanks a lot, I understand better :) It's creating a blastdb_p.loc file Could it be a similar problem with bwa ? I have installed the latest tool version I guess (https://toolshed.g2.bx.psu.edu/view/devteam/bwa/8d2a528a9513) which installed 2 tools "Map with BWA-MEM" and "Map with BWA". Should I get at some point a bwa_mem_index.loc ? but then "Map with BWA" tool should get access to indexes, right ?
The logic for data tables is that :
When you add items to the tables using biomaj2galaxy, the first part of each expression is the data table name (e.g. blastdb_p
in blastdb_p:blast/Anopheles_gambiae_protein:Anopheles gambiae protein
).
As some data tables name are not very intuitive, biomaj2galaxy automatically translates a few good-looking names to the real names (like bwa_mem
which is understood as bwa_mem_indexes
, see the code at https://github.com/genouest/biomaj2galaxy/blob/2001dc1284add9bc9b2193a70bc62a18dd1415bc/biomaj2galaxy/commands/add.py#L88)
Tell me if it's still not clear!
Thanks for the complementary information and your time.
For bwa indexes, I had to clean the shed_tool_data_table_conf.xml file to finally see the reference genome appearing in the bwa tools scroll-down menu.
Cheers,
Ok, great! Closing this, dno't hesitate to open another issue if you have other problems (I hope not!) Cheers
Hi, I am using latest BioMAJ and BioMAJ2 Galaxy versions and Galaxy v18.05. For example, Anopheles_gambiae.properties looks like:
bwa.loc file:
blastdb.loc file:
BioMAJ went through the update but bwa and blast indexes are not accessible through galaxy interface (bowtie2, star and Hisat2 are fine).
Would you have hints where to dig ... ? Thanks,