johnlees
commented
3 years ago I wouldn't take the listed heritability estimate as gospel, especially with a binary phenotype. I've seen it shunted to either end (0 or 1) when the true value should be somewhere in-between. You might want to have a look with another method which better models h^2 if you are interested in its value – the pyseer estimation is pretty 'rough and ready'
Running pyseer LMM on a set of unitigs derived from cholera samples, at some concentrations of antibiotics while heritability is measures as zero (h^2 = 0.00), there are significant hits in the results. I was wondering what can be the explanation of the results! can they be considered as real hits or they should be ignored as false positives? Thanks.