-
Notifications
You must be signed in to change notification settings - Fork 1
Description
Dear authors
I would like to utilize your precise deconvolution method, and confirm that I could correctly build the environment using P1.2 data in Nat Commun 2018. According to the article, I should run factor analysis on raw count not normalized data, but I could not obtain it.
Actually, I could obtain the dataset from SpatialDB; however, that does not include raw count data.
The log file of "Supplementary Dataset 1" in the Nat Commun article says "/home/emelie.berglund/count_data/1.2_10factors/P1.2.tsv" is the original file.
Would you please share the file with me?
I would really appreciate it if you help me!
P.S.:
The file "experiment0000-counts.tsv" is not the same as the raw count data, is it?
I tried deconvolving the file but the data deconvolved was not the same