Update MM tag regex to include upper case "Ambiguity codes"

[cmdcolin]

The regex for the MM tag (section 1.7):

  MM:Z:([ACGTUN][-+]([a-z]+|[0-9]+)[.?]?(,[0-9]+)*;)*

specifies ([a-z]+|[0-9]+) for the modification codes which is lower case only but this does not therefore parse the ambiguity codes like "C - Ambiguity code; any C mod"

the spec may want to update it to

([a-z]+|[A-Z]|[0-9]+)

[jkbonfield]

Edit: oops sorry I misremembered this work. We already have "C" as "any modified base", but it's not accepted by the regexp.

Good spot and thanks.

The question though is now whether we're permitted to describe the probability of e.g. "any modified C" as well as specifically 5mC and 5hmC. With the regexp above, we could describe both 5mC and 5hmC together, but not also a joint probability of "unspecified modification". However we could do a ;C+C,1,2,3... style modification to list it explicitly in a subsequent array of positions. That seems to imply there's little benefit to forbidding mixing C with m and/or h. Hence maybe MM:Z:([ACGTUN][-+]([a-zA-Z]+|[0-9]+)[.?]?(,[0-9]+)*;)* is the better regexp.

[cmdcolin]

yep, got the original message and edit. just for reference someone reported that modkit might have outputted this, was from a bug report i received here GMOD/jbrowse-components#5523 I haven't investigated modkit yet though