Closed abedkurdi closed 1 month ago
Hi @abedkurdi, sorry about the delay, I'm looking into this
Hello @sheridar The reason behind this is bootstrapping.
This issue seems resolved, feel free to open a new issue if you have any other issues/questions
Hello there,
First, I would like to thank you for this great package, I have been looking for such tool a while ago! Second, I was comparing the diversity plots, output from your package and output from
scRepertoire
R/Bioconductor package.I thought I would be able to reproduce, if not identical results, something very similar when using both packages. However, it was not the case
I counted the number of unique clonotypes in both analyses and I got the exact same number for each hashtag.
Do you have any idea what could be happening here? I appreciate your help.