The term "comma" by itself is somewhat ambiguous, since there are different kinds-- e.g., Pythagorean versus Ptolemaic.
I think "comma" without a qualifier is usually taken to mean a Ptolemaic ("ordinary" or "syntonic") comma, the interval between a "just major third" (5:4) and "Pythagorean major third" (81:64), which is equal to 81:80-- i.e., (81/64) / (5/4) = (81/64) * (4/5) = (81/64) * (64/80) = 81/80.
On the other hand, a Pythagorean ("diatonic") comma is the interval between twelve Pythagorean semitones (531441:262144, or [3^12]:[2^12]) and a perfect octave (2:1), which is equal to 531441:524288 (sometimes given as approximately 74:73).
In any case, a particular type of comma always has the same ratio, and an equal temperament semitone likewise always has the same ratio, so if we express an equal temperament semitone in terms of a comma then it will always be the same no matter which two adjacent notes we're considering-- C to C#, C# to D, D to D#, etc.
Since the two types of commas mentioned above are defined as the ratio between two different "versions" of a particular interval, by extension we could define other types of commas based on other intervals and whatever ratios we happen to be using to define them-- e.g., there is Archytas' comma. But as I understand it, a "comma" is relevant only in reference to some type of tuning system-- it's irrelevant in equal temperament, since by definition every equal temperament semitone is exactly equal to every other semitone-- 2^(1/12):1.