modkit pileup filter-threshold results in different calls

[mattloose]

Hi,

As I understand it modkit pileup defaults to calculating a threshold for calling methylation based on the input data.

In some circumstances this can lead to different methylation calls for the same read in a dataset. Even relatively small shifts in the filter-threshold can change methylation calls.

Is there a sensbile default that can be used in the general case?

If looking for reads with specific methylation patterns and they are fluctuating over the course of an analysis depending on data elsewhere in the genome it quickly becomes very complex!

Thanks.