Another analysis I did was gene-based. That means that instead of focusing on how many CNVs were present in each child, I first counted how many kids had gene G with a given CNV. For example, using the clean set of CNVs, we get a table like this (\ref{333900}):