Define omega per sample QCs #318
Description
The significance of the omega values per sample is hard to reach because of the sparsity of the data.
Despite this, in some cases we can still use non-significant values, but we need to have a sparsity metric that allows us to define whether we can or cannot trust the values of omega per sample.
In particular, we need to assess the sparsity in the number of truncating or missense mutations per sample.
For this we there might be different options, but a preferred one would include:
- Computing a value of mutation density per consequence type per gene. as described here:
- Compute the mutation density described above with plus and minus one mutation.
- With this, define a metric that measures the stability of this measurement.
- Set a threshold for this stability and apply it as a criteria for reporting the analysis or doing the regressions with clinical variables
to discuss between @koszulordie , @rochamorro1 and me (@FerriolCalvet )