levels program, frac_meth

[Sara Hosseini]

Hi,

I have a question about computation of fractional methylation level by "levels" program.

I get different fraction of methylated C when I used methpipe level program in comparison with binomial test, in which I consider the error rate based on lambda DNA (about 0.005).

The difference is noticeable. While level program estimates about 2 % of C are methylated, the binomial test shows me about 10 %.

I wonder how the levels program calculate the fraction of meC and if it considers any error rate while calculating this fraction and if yes what is that by default?

Thanks for your help.

Sara