Genomic Pathway Approach to Predict Brain Disorders

Principle Investigator: Demetrius M. Maraganore, M.D.

We developed a method to predict brain disorders. We first considered the axon guidance pathway (chemical signals that wire and repair the brain) and Parkinson disease (PD). We mined a Mayo genome-wide association dataset and identified single nucleotide polymorphisms (SNPs) within axon guidance pathway genes. We then constructed SNP models that predicted PD susceptibility (odds ratio=90.8, p=4.64×10-38), survival free of PD (hazards ratio=19.0, p=5.43×10-48), and age at onset of PD (R2=0.68, p=1.68×10-51). Mining of public genome-wide association and expression profiling datasets validated our findings. An independent analysis using different methods concluded that axon guidance was the pathway most significantly associated with PD in the Mayo dataset. We also constructed axon guidance pathway SNP models that were predictive of Amyotrophic Lateral Sclerosis (ALS); our final models for PD and ALS had sensitivities and specificities ≥92%. We will employ our genomic pathway approach: 1) to construct axon guidance pathway SNP models that predict Alzheimer disease, attention deficit hyperactivity disorder (ADHD), bipolar disorder, major depression, schizophrenia, and stroke; 2) to screen other pathways in PD, ALS, and these other brain disorders, and to identify additional candidate pathways; 3) to construct SNP models for additional candidate pathways that predict brain disorders (several new patents).