Functional predictions for UDN genes

UDN is researching rare genetic diseases and listed the genes associated with such diseases.

Since our lab developed several methods for predicting gene function we would like to ask: Can we predict the phenotype for these UDN genes? If we are able to do it, which diseases are matching our predictions?

Determine which co-expression prediction matrix had the most matches with ontology terms with HPO and MGI. Do all the prediction matches produce the same # of hits or is one method significantly better than the rest? Is the same prediction method better for both HGO and MGI libraries or are they different?

Does the ranks of the hits suggest that the prediction models should be improved to improve phenotype hits and ranks?

If matches were found for each prediction method, were the phenotypes found the same?

Does the UDN genes share the same common phenotypes and if so, is there a genetic link amongst all these shared common phenotypes? Can we predict a drug to be tested in cell based assays based on this genetic link?