For first-time founders, ignorance of the immense challenges ahead can be beneficial. This "naivete" prevents paralysis, enabling them to dive in, learn from direct experience, and adapt quickly without being overwhelmed by the full scope of the difficulties from the start.
Before the DSM, psychiatric disease definitions were bespoke to different schools of thought. The DSM's great achievement was creating a common diagnostic umbrella. While imperfect, this standardization was a crucial foundational step for the field to begin communicating and researching consistently.
The initial goal of precision psychiatry isn't complex machine learning or perfect biomarkers. It's about systematically collecting basic, meaningful data—like cognitive function—that we already know correlates with treatment outcomes. This simple act of consistent measurement provides a powerful foundation for better understanding patients.
Instead of a traditional biomarker-first approach, Alto took a drug already approved in Europe (agomelatine) and used machine learning on patient EEG data to discover a novel biomarker that predicts who responds best. This "reverse-engineering" approach de-risks development by finding a precise population for a proven drug.
Despite hype around genetics and multi-omics, their value in stratifying psychiatric patients is limited. Genetic variants have small effects, and peripheral samples like blood poorly reflect brain biology (e.g., blood serotonin comes from platelets, not the brain). Direct brain function measures are more reliable.
A patient's subjective report on their cognitive ability correlates more strongly with their overall mood than with objective cognitive test results. This disconnect reveals why objective measures like EEG or behavioral tests are essential; self-perception is an unreliable proxy for the underlying biological processes that need treatment.
Decades-old symptom scales are often criticized but are irreplaceable because they possess "face validity"—they measure the symptoms patients actually experience. While they lack mechanistic insight, they capture the patient's subjective reality, which is the ultimate endpoint of any psychiatric treatment. No objective measure can replace asking "do they feel better?"
The adoption of precision medicine in psychiatry will mirror oncology's journey. It won't happen overnight. The field first needs an initial, landmark success with a targeted therapy (its "Herceptin moment") to shift mindsets and standardize data collection. Only then can it progress to a full-blown revolution where precision is the norm.
The initial hurdle for precision psychiatry isn't achieving 100% accuracy. The goal is to be meaningfully better than the current trial-and-error standard. Moving the needle on treatment remission from 30% to 40-45% would be a huge clinical success, creating a new benchmark and starting a virtuous cycle of improvement.
Traditional boundaries between neurology and psychiatry are artificial. From a brain circuit perspective, conditions overlap significantly—Parkinson's involves mood and cognition, and depression involves key neurological pathways. The focus should be on matching a measurable circuit dysfunction with a targeted therapy, regardless of the clinical specialty's historical label.