search

aertslab / AUCell

AUCell: score single cells with gene regulatory networks
110 stars 26 forks source link

Recommended for use on Aggregated data? #4

Closed daacarri closed 5 years ago

daacarri commented 5 years ago

Hello,

I'm trying to find cells with a hypoxic signature, but am using aggregated data from two different experimental conditions.

I am getting very low numbers, and was wondering if it isn't because I'm using the aggregated data. Would it be better to identify cells from each of the individual data sets?

Any help would be appreciated.

s-aibar commented 5 years ago

Hello,

I'm not sure what do you mean with "aggregated data" (do you mean concatenated datasets?). In principle, it is fine to just cbind the expression matrices or rankings if they come from similar datasets (e.g. so if there is similar coverage; the ranking is calculated individually for each cell).

However, to understand what is going on you might need to check it a bit further (maybe you need to adjust the aucMaxRank?).