Closed sparthib closed 2 years ago
I think that we could also use dynamic thresholds for PpTau
and PAbeta
(same formulas as for n
).
Then we might consider a spot to have pTau
pathology in #25 if it's either an outlier on PpTau
or NpTau
(similar for Abeta
).
Based on Br_3874 in https://github.com/LieberInstitute/Visium_IF_AD/tree/master/plots/07_spot_qc/outliers
NpTau = max(NpTau of 004 and 106) same with PpTau NAbeta = max(Q2 for all controls) or max(Q3 for all controls)