Closed maximilianh closed 1 year ago
closing this. I reached out to Bob Jones and the number wasn't 2/3 but rather 20% and there are very good explanations why some cells were filtered (only ERCCs, low quality cells)
So 20% of the BAM files that are not used in the H5AD are entirely expected.
Is there an explanation somewhere why the cell-type annotation in the H5AD for the Smartseq2 data are missing 2/3 of the cells? Thanks! Max