The 750,000-individual study suggests that our common diagnostic criteria may need substantial revision.

This week I've been reading this paper, appearing in Cell, describing a truly herculean effort to determine some of the genetic underpinnings of psychiatric disorders.

The study begins as a huge genome-wide association study, but man, it has layers. Like an ogre.

Let's start at the top and dig our way in.

Researchers obtained genome data from 232,964 individuals with psychiatric disorders. To put that in context, that's roughly 2 New Haven's worth of people. The vast majority had major depression, but multiple other psychiatric disorders, as you can see here, were present.

They also had 4 New Haven's worth of controls - nearly 500,000 individuals with genome data. It's crazy big. Though this population wasn't as diverse as real New Haven - all individuals were of self-identified European descent.

But those numbers allowed the researchers to search across the entire genome for small variations in genetic code that were seen much more frequently in those with psychiatric disorders.

All told, they found 136 such hotspots in the genome - that's all the dots above the red dashed line in this Manhattan plot here.

That included 35 never-before reported hotspots, which is pretty incredible. And they could have stopped there, but this is really just the surface of the paper.

The researchers were on the hunt for so-called "pleotropic loci" - hotspots that didn't just appear to confer risk for one psychiatric disorder, but for multiple. The idea is that these hotspots would help us begin to understand if there are common processes central to all psychiatric disease, and, of course, develop more universal treatments.

109 hotspots were pleotropic - linked to more than one disorder. Those genomic connections allowed researchers to create this network map which reveals how these diagnoses are genetically linked.

SCZ: Schizophrenia. BIP: Bipolar Disorder. AN: Anorexia Nervosa. OCD: Obsessive-Compulsive Disorder. MD: Major Depression. ASD: Autism Spectrum Disorder. TS: Tourette's SyndromeWhat I find so cool here is how closely this genetically-derived map matches what we observe clinically. Bipolar disorder and schizophrenia are strongly linked. As are anorexia nervosa and obsessive-compulsive disorder. The link between autism spectrum disorder and ADHD is no surprise, but some novel findings bear more research - like the genetic link between autism and major depression.

In fact, several different bioinformatic techniques revealed the pattern you see here: three broad groups of disorders, likely representing the sequela of related genetic processes. Someday, this may redefine how we classify psychiatric disease.

Now I mentioned the hunt for pleiotropy. Well - one hotspot stood out in this manner above all the rest - a small mutation in a gene called DCC (of colon cancer fame) - it was associated with all 8 psychiatric diagnoses in the dataset.

This gene is much more than "deleted in colon cancer" though - it guides axonal growth in neuro-development.

Germline loss-of-function mutations in DCC cause severe neurodevelopmental syndromes, and are often embryonic lethal - we're not talking about a non-functioning gene here. Just one that is functioning slightly differently - enough to potentially create a brain that is more susceptible to the environmental triggers of psychiatric disorders.

Are drugs targeting DCC going to rid the world of psychiatric disease? Of course not, but the understanding we gain from genetic studies may well redefine how we think of psychiatric disease, and though that may panic the editors of DSM-VI, it may benefit our patients in the end.

This commentary first appeared on medscape.com.