For files of the <GEO>_signals.txt.gz variety, it won't apply p-value detection to probes if the p-value column is there for each sample.
IF meth and unmeth column names differ in case (e.g. SampleA Unmethylated Signal vs samplea Methylated Signal) it wont work. Instead you have to use pandas like this:
>>>import pandas as pd
>>> x = pd.read_csv('GSE67530_signals.txt', sep='\t')
>>> cols = {i:i.title() for i in list(x.columns)}
>>> x = x.rename(columns=cols)
>>> x.to_csv('GSE67530_signals_revised.txt', sep='\t')
>>> z = methylprep.read_geo('GSE67530_signals_revised.txt')
>>> z.to_csv('GSE67530_betas.csv.gz')
(make samples case insensitive. It's possible to have the function try this trick before failing)
<GEO>_signals.txt.gzvariety, it won't apply p-value detection to probes if the p-value column is there for each sample.(make samples case insensitive. It's possible to have the function try this trick before failing)