Predicting whole genome features whole exome data?

Question

[unksci]

Are there methods or computational tools to predict whole genome features from whole exome data alone? Which types of features could be predicted to which certainty?

The rationale of my questions stems from whole exome data being more abundant, and common characteristics relating to linkage disequilibrium and evolutionary events, among other factors, surfacing in exome data as well as whole genome data.

[Kevin Blighe]

Please elaborate on what you mean by 'features'?

[unksci]

Could be something like CNVs