dpeerlab/Palantir

input data and imputation

wangjiawen2013 opened this issue · 1 comments

Hi,
what kind of data should be used as input when running palantir ? normalized data or scaled data
I use scanpy.external.tl.palantir (https://scanpy.readthedocs.io/en/stable/generated/scanpy.external.tl.palantir.html#scanpy.external.tl.palantir) for trajectory inference, but I dont know whether to scale the data or not. Does MAGIC imputation also work on scaled data ?

We recommend the use of normalized data and not scaled data. We have not test Palantir on scaled data but if scaling is necessary for your data analysis, you can try running Palantir on scaled data as well. MAGIC imputation is not as well defined with scaled data