The main pending problem would be how to estimate the background indel rate? I envisage two avenues: 1) To what extent indel rate is associated with synonymous mutation rate? Regress indel counts (recurrent or unique) against synonymous 2) Fit model with covariates along non-driver genes, this is related with #15 Once the background indel rate is estimated, we can proceed along the same principles of Omega for SNVs: * Poisson or Negative Binomial count model depending on the background mutation rate and dN/dS parameter * Compute dN/dS confidence bounds assuming chi-squared-distributed likelihood-ratios
The main pending problem would be how to estimate the background indel rate? I envisage two avenues:
Once the background indel rate is estimated, we can proceed along the same principles of Omega for SNVs: