search

satijalab / sctransform

R package for modeling single cell UMI expression data using regularized negative binomial regression
GNU General Public License v3.0
203 stars 33 forks source link

Questions about finding different gene-peak links based on rna and sct (related to sctransform) #156

Closed HelloWorldLTY closed 1 year ago

HelloWorldLTY commented 1 year ago

Hi all, I contact two seperated datasets to generate the overall gene-peak link result. However, I found that the results I have based on RNA in the idenfitifcation process are very differernt from the results based on SCT. For two datasets, is it ok for me to directly contact them? Thanks.

saketkc commented 1 year ago

The gene-peak link functionality is part of Signac.

Based on whoch assay you chose the links can vary across the two assays because of the different nature of normalization (LogNorm vs SCT). You would want to stick to one method across the datasets.

HelloWorldLTY commented 1 year ago

Hi, for all the datasets I used, they follow the pipeline of SCT. For all the datasets we used to detect the link, we used SCT. Therefore, my question is , if I combine two datasets processed by SCT to one array, the number of links I found is quite large and different from the number of links I found based on LogNorm. Thanks.