I recently developed a method for specifying a comprehensive gene list for investigating genes related to psychiatric violence, which I describe below. First though, here’s a cool picture from the analysis:
I started by extracting a list of diseases involving violence from , removing epilepsy, dementia, mental retardation (is there a better word for this?), and Alzheimer’s disease. Also removed sexual sadism from the list as one might debate whether or not this qualifies as “disease”. I then matched those diseases by name–more accurately by components within each name–to diseases contained in DisGeNET  to determine genes associated with those diseases. Next, I built a network graph of the genes where an edge between two genes indicates one or more diseases in common. A tractable subset of this graph is pictured above for demonstration. Finally, I computed the size of the ego graph for each gene (node) and ranked them, as listed below. Greater ego network size indicates greater probable biological importance vis-a-vis psychiatric diseases having violence as a symptom.
I posted my code used for this analysis at .
Here is a portion of the gene list I put together:
|Ego Network Size||NCBI Gene Symbol||NCBI Gene ID|
|. . .||. . .||. . .|