what's the meaning of "WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig(contig_name). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs." in CNV Calling?
I use Docker GATK 4.2.0.0.
I want to detect CNV in Saccharomyces cerevisiae
I use the following command
gatk DetermineGermlineContigPloidy
-L cohort.gc.filtered.interval_list \
--interval-merging-rule OVERLAPPING_ONLY \
-I XXX.counts.hdf5
-I ...
--contig-ploidy-priors contig_ploidy_priors.tsv \
--output ../results \
--output-prefix ploidy \
--verbosity DEBUG
The content of contig_ploidy_priors.tsv is (the last line is mitochondria):
CONTIG_NAME PLOIDY_PRIOR_0 PLOIDY_PRIOR_1 PLOIDY_PRIOR_2 PLOIDY_PRIOR_3 PLOIDY_PRIOR_4
NC_001133.9 0.01 0.01 0.96 0.01 0.01
NC_001134.8 0.01 0.01 0.96 0.01 0.01
...
NC_001224.1 0.97 0.01 0.01 0.005 0.005
There are no errors in the output report, but every sample, every chromosome has the following warning, such as:
10:59:06.863 WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig (NC_001133.9). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs.
10:59:06.863 WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig (NC_001134.8). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs.
10:59:06.863 WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig (NC_001135.5). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs.
10:59:06.863 WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig (NC_001136.10). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs.
10:59:06.863 WARNING gcnvkernel.structs.metadata - Sample XXX has an unrecognized contig (NC_001137.3). Germline contig ploidy determination may not be reliable for decoy/non-standard contigs.…………
What's the meaning of these warnings? Should I ignore these warnings?
In addition, should I exclude mitochondrial information from CNV calling?
-
Hi rq m,
The gCNV tools were written with human samples in mind, so we haven't tested it for your use case. However, it is possible to run these tools on other species.
These warning messages indicate that you should check the ploidy results in the contig_ploidy.tsv inside the contig-ploidy-calls directory. You can determine there if the results are reliable for the unrecognized contigs.
And yes, you should exclude the mitochondrial information from your CNV calling.
Best,
Genevieve
Please sign in to leave a comment.
1 comment