meren
commented
1 month ago Hi @NebulaSL,
This is interesting. But the problem you run into, i.e., "profile.db does not have a table named mean_coverage_Q2Q3_splits", is not related to --min-mean-coverage or --min-coverage-for-variability flags. I wish you had not gone that direction :)
The best practice for refining external binning results is to create a single contigs-db using all contigs, and then profile your metagenomes, and import your final collection into the final merged profile database. But that is another topic.
The only reason I can think of why a profile database will be missing a mean_coverage_Q2Q3_splits table is the following: something went wrong and your process was killed prematurely. If the last message you see in your log is the following,
====================================================
According to the data generated in the contigs database, there are 2117978
contigs in your BAM file with 0 gene calls. Which may not be unusual if (a) some
of your contigs are very short, or (b) your the gene caller was not capable of
dealing with the type of data you had. If you would like to take a look
yourself, here is one contig that is missing any genes:
NODE_922990_length_375_cov_1.446309"it means the profiling task certainly did not finish. You should this line at the end of the log for a successful profiling run:
* Happy πIndeed, this is what I see when I run the following command on your profile-db:
sqlite3 PROFILE.db '.tables'
collections_bins_info item_additional_data states
collections_info item_orders variable_codons
collections_of_contigs layer_additional_data variable_nucleotides
collections_of_splits layer_orders views
indels selfIn contrast to this output, which is coming from a profile-db that is successfully finished:
sqlite3 PROFILE.db '.tables'
abundance_contigs mean_coverage_Q2Q3_splits
abundance_splits mean_coverage_contigs
collections_bins_info mean_coverage_splits
collections_info metabolite_abundances
collections_of_contigs protein_abundances
collections_of_splits self
detection_contigs states
detection_splits std_coverage_contigs
indels std_coverage_splits
item_additional_data variability_contigs
item_orders variability_splits
layer_additional_data variable_codons
layer_orders variable_nucleotides
mean_coverage_Q2Q3_contigs viewsPlease re-create your profile db (without using --min-mean-coverage, --min-coverage-for-variability, or --report-variability-full unless you have other reasons to believe that you need them), and see if there are any server side / HPC problems.
Best wishes, Meren
NebulaSL
Short description of the problem
Hello Anvio team!
I am trying to use anvio for automated bin refinement (which is generated outside anvio). I created contig db for each bin and profile db (using the bam file mapping from my reads to assembled contig). Then I tried anvi-import-collection, however it gave me an error said my profile.db does not have a table named
mean_coverage_Q2Q3_splits. I saw the post in [https://github.com/merenlab/anvio/issues/1148] so I went back on my profile.db, and tried the similar solution in the post using --min-mean-coverage 0 during profile creating, however I got an error as anvi-profile: error: unrecognized arguments: --min-mean-coverage 0. Then I also tried --min-coverage-for-variability 0, problem is still there.related code and output are as followed: error for import collection:
how my profile db look like:
code for generate the profile db:
log file for profile generatingοΌ
anvi'o version
System info
I am using the linux server provided by the univ. I did not install anvio by myself.
Files / commands to reproduce the issue
a contigs database, a profile database and BAM file are provied in via the link below
https://drive.google.com/file/d/1uTWNw-BBW5DC5b5_06ito6H5BOj0o7Ni/view?usp=sharing
If you don't have files to share, please remove this section along with the header.
Thanks in advance!