Incorrect n_mod value in minimod freq bedmethyl output

[psummanen]

Hi!

There seems to be an error in the n_mod value in minimod freq bedmethyl output.

Below is freq output in both tsv and bedmethyl formats. The n_mod column in bedmethyl seems to show the n_called value, not the n_mod value. Freq value is consistent in both formats.

Output from minimod freq in tsv format:

contig          start   end     strand  n_called    n_mod   freq            mod_code
Escherichia     381     381     +       235         7       0.029787        a
Escherichia     381     381     -       206         3       0.014563        a

Output from minimod freq in bedmethyl format:

contig            start   end    mod_code   n_mod    strand       freq
Escherichia       381     382    a          235      +            2.978723
Escherichia       381     382    a          206      -            1.456311

[imsuneth]

Hi @psummanen,

Thanks for opening the issue.

The TSV output looks correct. But, did you run minimod freq with -b to get that bedmethyl output?

The usual minimod freq bedmethyl should look like this as documented here

chr22	20016387	20016388	m	4	-	20016387	20016388	255,0,0	4	0.000000
chr22	20016820	20016821	m	1	+	20016820	20016821	255,0,0	1	0.000000

[psummanen]

Hi!

Yes, I used the -b option in minimod freq to get output in bedmethyl format. In the example above I had removed columns 7, 8, 9 and 10 from the output. Below is an example of the bedmethyl output with all columns.

Escherichia     381     382     a       235     +       381     382     255,0,0 235     2.978723
Escherichia     381     382     a       206     -       381     382     255,0,0 206     1.456311

I used the latest 0.5.0 version of minimod to get the output.

[imsuneth]

Hi @psummanen,

It is a mistake in the minimod README.md. Minimod freq bedmethyl outputs both the 5th and 10th columns equal to n_called. n_mod is n_called*freq/100. Thank you for pointing it out. It's fixed in the documentation now.