Closed JSchapke closed 4 years ago
In the normalize_per_cell function, you normalize the matrix by rows and then multiply the results by 1e6. What is the purpose of the multiplication?
This is essentially the idea behind the TPM (transcript per millions). It is better to have larger numbers to avoid machine precision errors propagating in downstream analysis.
In the normalize_per_cell function, you normalize the matrix by rows and then multiply the results by 1e6. What is the purpose of the multiplication?