CheckM icon indicating copy to clipboard operation
CheckM copied to clipboard

QA failed not placed in tree - root

Open raw937 opened this issue 3 years ago • 5 comments

Good evening,

I have a single circular contig from a cyanobacteria. When I run the standard: checkm lineage_wf -t 36 -x fasta -f genome_checkm.tab --tab_table /data/folder /data/folder/checkm

I get no markers found, no calculations, no placement in the tree. It is assembled from nanopore but the headers look fine etc. I can pull full length 16S, 23S, and RecA etc. I just trying to get the completeness/contamination stats.

Ran a close relative from NCBI in the same run and it worked fine.

Lost on why this would be?

many thanks, Rick

raw937 avatar Sep 14 '22 22:09 raw937

I have tested on checkm on kbase similar issue. It appears to have issues with identification of the marker genes. I ran prokka can could pull them all out?

Any thoughts?

raw937 avatar Sep 14 '22 22:09 raw937

Not sure. Does the gene calling look correct (i.e. expected number of genes of the expected length)? My best guess is that it is an assembly error that is resulting in poor gene calling and thus CheckM (really HMMER) is unable to identify the marker genes.

donovan-h-parks avatar Sep 15 '22 00:09 donovan-h-parks

Its so weird. As when I run prokka get the faa and run checkm I get 50% complete at 3 Mbp genome. Similar to its reference. Its one circular contig.

raw937 avatar Sep 15 '22 00:09 raw937

Could this be due to insertions and deletions causing frameshift errors? This was a known issue with Nanopore sequencing that has only recently been addressed: https://www.nature.com/articles/s41592-022-01539-7

donovan-h-parks avatar Sep 17 '22 01:09 donovan-h-parks

It appears to be true. Just new to it. Just curious if you have seen this before ?

raw937 avatar Sep 17 '22 02:09 raw937