One of the most challenging issues in studying mental health during COVID-19 was separating the impact of containment policies (lockdowns) from pandemic severity (societal infection rates), as both can contribute to mental distress. In this extensive international effort, we aimed to address this 🧵

Despite D.Phil. training in AI and neuroscience, I’m excited to share my first manuscript in human genetics (now published in AJHG)! Our results show that relative genetic risk varies substantially across sociodemographic contexts (see comments). www.cell.com/ajhg/fulltex...

Some slides from a recent talk on missing heritability. www.dropbox.com/scl/fi/kvogj...

Can LLMs infer depression severity from clinical notes? Here, we evaluate GPT-5.2 on outpatient psychiatry notes to infer depression severity. Results show moderate agreement with PHQ-9 and stronger agreement with clinician chart review (see comments). www.medrxiv.org/content/10.6...

I came across this Nature paper co-authored by a former Oxford colleague. I think it’s interesting for two reasons (see comments). www.nature.com/articles/s41...