control over cell/gene filtering

[andynkili]

@AntonioDeFalco ,

Thank you for your helpful and user-friendly tool.
The current approach for cells and genes filtering ("cells with less than 200 detected genes and the genes expressed in less than 1% of cells are removed") might be too 'general' especially for genes differentially expressed in small clusters: in our case, some of our datasets include very small quiescent cell type population (~ 100 cells not expressing many genes, but the latter are differentially expressed).
In that particular case, removing genes not expressed in at least 1% of the whole dataset and cells not expressing at least 200 genes may lead to missing relevant CNVs.
Would it be possible to turn off filtering or better: give more control over the parameter threshold so we can manually draw the line between noise and actual CNVs?

Best,
Andy

[AntonioDeFalco]

Hi @andynkili,
If you install the last version of SCEVAN in the pipelineCNA function, you can now use the parameter perc_genes and use your percentage. However, I think it is impossible to analyze cells with less than 200 genes; I consider them to be of too low quality and unusable for copy number analysis. But if you want to try, you can modify the preprocessingMtx function in the GitHub.

Thanks. Regards