See, this calculation here:
(Potency = Assay purity * active fraction * (1-water content) )
I don't really understand. It may be a shorthand for when you know all the data needed and you just plug it in. It seems like you don't, so maybe we can skip it for now.
OK, lets try to work through this stepwise:
Minocycline hydrochloride, crystalline:
This helps determine molarity or a solution. Not using molarity, skip it.
water by Karl Fischer: < OR = 10% (3MOL/MOL)
This is how much water was in it when they tested it. This is not a very exact number. Remember that. This is not a by mass ratio. Remember that too.
Purity by TLC: > OR = 98%, Elemental Analysis (anhydrous basis): 54.5 TO 57.3% Carbon 8.2 TO 8.8% Nitrogen,
Useless info, IMHO, because ...
Potency (anhydrous basis): 890 TO 950 micrograms minocycline/mg
This is the interesting bit. Every mg of powder from this vial, if it were perfectly dry, will act as if it were somewhere between 0.89 and 0.95 mg of minocycline. What is the rest? I don't know. Perhaps random compounds, perhaps derivatives with similar activity but weaker because of some structural change or bound inhibitor or something like that.
First off, you have to determine how much water there is. You have to do a Karl Fisher titration, and get a result with sufficient precision for your needs. Maybe you can dry it first, the USP (and friends) will tell you how to do that. You might dry under vacuum, with or without heat, for a certain period of time -- hours or days. This might be enough to remove all moisture, and you won't need to correct for moisture, or maybe not. In that case, you'll determine how much more power you'll weigh out, to compensate.
Then you'll have to, as an example, pick the middle value for activity, for example, 0.92 mg activity/mg powder and calculate ... what is the resulting solution if the activity is actually 0.95? What is the activity if it is really 0.89? For that matter, what if it is 10% water. Maybe the results show your activity isn't as important.