Major depressive disorder (MDD) carries a 1 in 6 life-time risk, has the highest morbidity of any mental health condition, plus very substantial social, welfare and economic costs. In Europe alone the estimated direct costs are €42 billion/annum, with indirect costs of an additional €76 billion/ annum. MDD is poorly treated: at best 30% of patients respond well to any chosen course of therapy, while 30% fail to respond to any of the currently available treatments. New approaches to early intervention and prevention of MDD and associated co-morbidities are urgently needed for reducing these costs. We will use machine learning techniques proven in other contexts, and a newly available Generation Scotland cohort, to develop personalised algorithms predictive of MDD and associated co-morbidities. If proven, this will facilitate further investment in precision diagnostics for healthcare and pharmaceuticals. Our primary aim is to improve prediction to target early and optimal intervention.