Requirements for the input data

[Epoch1128]

Thanks for the good tools. My issue is as follow. If I'd like to get the cell proportion in each spot based on Spatial Transcriptomics data and Sc-RNA reference what should I do to satisfy the model input. E.g. can I just use the raw count data in ST and ScRNA dataset or I need to do normalization and log() operation before mapping cause I noticed you have mentioned in your paper "we do not assume that G and S measure gene expression using the same unit of measures" in Methods part. Does it mean the raw count data is OK? Thanks!

[Hejin0701]

Hi @Epoch1128 . It will still be better if you normalize the data before mapping. But no need to take log operation because Tangram linearly combines the gene expression between different cells.