jodyphelan / tbdb

Standard database for the TBProfiler tool
GNU Lesser General Public License v3.0
28 stars 18 forks source link

Remove two ahpC mutations #50

Closed jodyphelan closed 1 year ago

jodyphelan commented 1 year ago

Lack of evidence for Asp73His and Glu76Lys has been highlighted. The original source for these mutations was the TBDreaMDB website which listed drug resistance mutations but has since stopped operating. These mutations were investigated with a dataset of 50,722 samples for which Illumina WGS was available.

Asp73His

All of the samples with this mutation are lineage1.2.2.2 strains (86/86). A phylogenetic tree with the branch length representing the number of mutations occuring on each branch seems to indicate that this mutation is a phylogenetic mutation that would have predated the introduction of isoniazid.

Asp73His

Glu76Lys

The majority of the samples with this mutation are lineage1.2.2.2 strains (61/66). A phylogenetic tree with the branch length representing the number of mutations occuring on each branch seems to indicate that this mutation is a phylogenetic mutation that would have predated the introduction of isoniazid.

Glu76Lys

francesccoll commented 1 year ago

These two mutations (Asp73His and Glu76Lys) were originally extracted from TBDreaMDB. They were not identified as lineage-specific markers in the original dataset (https://doi.org/10.1186/s13073-015-0164-0) and therefore were kept in the final catalogue. The latest evidence (with a much larger dataset) suggests they are specific to certain sub-lineages, which makes them unlikely to be involved, or compensate for, isoniazid resistance.