A lung-specific cis-expression quantitative trait loci (cis-eQTLs) from GTEx v.820 (n = 515) and the Lung eQTL Consortium21 (n = 1,103) provided further support for a subset of loci (Supplementary Table 7), including FOXP4 (chr. 6p21.1) and ABO (chr. 9q34.2), OAS1/OAS3/OAS2 (chr. 12q24.13) and IFNAR2/IL10RB (21q22.11), where the COVID-19-associated variants modify gene expression in lung.
chr6:41,515,652G>C, GRCh38, rs9367106 Is the lead variant in the association study, but that doesn’t mean that it’s the causal variant. Genome Wide Association Studies line the one described here are relatively blunt instruments that are hard to interpret on their own, but can point towards next steps. Some plausible hypotheses are:
1) one or several mutations in the protein FOXP4 affects susceptibility to long covid [edit: reading the paper more after my coffee, the affected region isn’t actually in the coding region of FOXP4. “ no variant in LD with the lead variant is coding”]
2) a variant in a region that controls expression of FOXP4 affects long covid.
3) variant(s) in a gene near FOXP4 is actually what’s causing long covid (though this is somewhat less likely given the locations of the snps that are associated).
To make this more complicated, genes starting with FOX are usually transcription factors, meaning they control the expression of many other genes. So it’s certainly plausible that it’s a hinge point that controls a large, amorphous downstream response. I don’t know much about the biology of this gene in particular.
What one could say with moderate confidence after a 23andMe style test (assuming they have one or more associated variants) is that you might be at elevated risk for long covid, but only about twice the likelihood as the general population, assuming I’m reading the figures right. Not insignificant risk, but not perfectly predictive either. Where this result is more useful is directing further studies on the mechanism of long covid and (once we understand that) potentially ways to treat the underlying cause.
