Hi,
In another issue on this page you mention how mutational burden should be calculated:
if you want to get the callable sites with at least 10 cells, you should sum the column NC for all rows in the cell type with Cov >= 10.
In our manuscript, we computed the mutation load at cell-type resolution by using a minimum Cov >= 5 and the next formula:
(# somatic mutations in the cell type Z) / (# callable sites in the cell type Z)
I did this in my own work, but then I then I attempted to factor in the number of cells I had for each cell type by further dividing this value by the number of cells to get mutations per callable site per cell. (I then multiplied by 1 billion in order to get numbers that made more sense to people so I ended with mutations per callable site per 1 billion cells).
Is this a valid way of doing things? Or would it be better to stick with the value produced from the original equation?
Hi, In another issue on this page you mention how mutational burden should be calculated:
I did this in my own work, but then I then I attempted to factor in the number of cells I had for each cell type by further dividing this value by the number of cells to get mutations per callable site per cell. (I then multiplied by 1 billion in order to get numbers that made more sense to people so I ended with mutations per callable site per 1 billion cells).
Is this a valid way of doing things? Or would it be better to stick with the value produced from the original equation?
Thanks, Jake