Contamination in Pseudomonas sp000955805

GTDB
1

Hi, just want to report that there is a contamination in the genome GTDB - Loading.... The genomes has perfect matches to 16 different RNA sequences in the ERCC RNA spike-in mix. Considering the mix is artificially designed for quantification in RNAseq and therefore shouldn’t have any matches, let alone 16 different perfect matches in any bacteria, this is almost certainly a contamination issue. I came across this when analyzing the bacterial content of a RNAseq sample with ERCC spike-in.

I know that you guys are not responsible for cleaning up genomes, but maybe consider not using this genome as a representative species/strain, or not including it in your next release at all?

Thanks.

2

Thanks for the heads up, we’ll keep an eye out for this one in the next release

3

Hi zxl124,

Have much contamination do these 16 different RNA sequences represent? We do accept genomes with small amount of contamination into GTDB so long as they pass out QC criteria:

We have good evidence that contamination does impact taxonomic classification so have generally leaned towards capturing more biodiversity than aiming to have only contamination-free genomes:

That said, if the contamination here is substantial (>5% of bps) we can explicitly flag this as a genome that should fail QC.

Thanks,
Donovan