Improve expression level interchange model
david4096 opened this issue · 0 comments
Sometimes one receives an RNA expression table with genes and samples connected by some expression level. In the case when the units are not known, one can specify this using the ExpressionUnit enum.
However, the normalization may also be unknown. This issue is to change that boolean to an enum to allow for the case when normalization is not known. This will make it easy to correctly interchange expression level data when not all the metadata are available. @saupchurch
The need for this feature arose when trying to find sane defaults when these data are expected to be interchanged. Assuming normalization can dramatically alter the outcome of an analysis.
We should also provide a way to describe both the effective_count for multi-reads in addition to the raw_read_count so that new normalizations can be made.