This is a very broad question. The most frequent uses of WGS data provide information about the somatic changes that occur across the genome. Using WGS, you can identify mutations, copy number changes and structural variations that relate to a state of disease as long as you have a healthy control to compare to. You also can perform SNP profiling to look for inherited mutations across family samples and perform statistical analyses to determine risk for certain diseases.
WGS gives detailed data on the entire genome, so where exome sequencing (sequencing of only the coding regions) may cause you to miss certain genomic changes, WGS can allow you to capture this information in a more extensive context. WGS also allows you to identify novel regions of interest, such as long non coding RNAs, novel promoter sequences, and very large rearrangements.
What WGS won't give you is the gene expression information for the sample. To do this, you'd have to perform transcriptome sequencing.
Choosing WGS as your NGS strategy depends completely on what kinds of questions you are trying to answer.
There infinite number of possibilities you can find in data, but WGS project is not designed in a way to sequence and later decide what you can find in that. But you have to make hypothesis or any frame work what you would expect in the data. In this regard, this question is an ill framed question.