My story

short version   long version  

I first encountered the term "personalized medicine" around 2007. After studying genetics in undergrad I went to work for a small pharma called Vanda Pharmaceuticals. The CEO was the former head of pharmacogenomics at Novartis, and so in parallel to developing a drug for schizophrenia, we were also developing a panel of genetic markers to predict which patients it would work best for. (Unfortunately, the FDA mistakenly rejected the NDA for this drug. (They approved it eight months later after an appeal, but because of $$$, half the company had to be let go in that time. And so the whole framework for developing a companion diagnostic was lost)).

In the meantime, I went off to do a PhD in neuroscience, since I'd become fascinated with understanding how the brain works. I ended up doing my thesis on experience-dependent plasticity, because I liked the idea that the truly important stuff, what actually makes us who we are, isn’t the genetics (which is more like a blueprint), but the impact of experience on how our brains are wired.

It just so happened that the main finding of my research was that during development, the brain essentially learns by sending information backward through circuits to refine upstream connections. I discovered this right around when the three "godfathers of AI" wrote this review in Nature talking about how training artificial neural networks with this principle was allowing computers to see (and now just a decade later, think?).

So, after grad school, I did a fellowship in data science and went to work on human health data, excited to apply these tools to understand and improve human health. I spent two years at a fantastic company called PatientsLikeMe, and then five years at a nonprofit focused on advancing PTSD research, primarily working to understand how we might better differentiate or subtype psychiatric conditions to improve treatment outcomes. And over that time, I watched loads of resources get poured into measuring the biology of these conditions, all the while growing more and more convinced that a better way to approach it would be to focus first on the psychology. Something like depression can mean 100 different things to 100 different people, and by filtering through structured clinical interviews and the medical field's attempt at consensus, we're losing sight of what could easily be a dozen different "types" of depression.

I also think mental health is at a point in the public consciousness that it’s never been before. People are more willing to talk openly about it. The idea of crowd-sourcing the collection of this data is now possible in a way it wouldn’t have been ten years ago. (Plus, research will continue to make progress towards quantifying how our brains work, and it’s likely something like a portable EEG device will eventually get to the point where collecting that data at scale will be within reach).

So, I set up this site to start testing this idea, and to (slowly) try to build a dataset that could help toward this goal. (For clarity - the "team" is me and a few friends / ex-colleagues I've been bouncing ideas off).

Interested in helping? Thoughts/feedback?
Or give your email if you want to go on the mailing list: