
🩺🧠 ✨ Excited about our latest work published in @eBioMedicine! In the BioFIND dataset, we found that brain activity provided unique information about progression to Alzheimer’s Disease (AD). Paper: thelancet.com/journals/ebiom…; Thread 🧶 ⤵️

Neuroscientist in Pharma R&D @Roche. Passionate about brains, machine learning, metabolism, health, nutrition ... cycling. Tweets = personal views & opinions.

🩺🧠 ✨ Excited about our latest work published in @eBioMedicine! In the BioFIND dataset, we found that brain activity provided unique information about progression to Alzheimer’s Disease (AD). Paper: thelancet.com/journals/ebiom…; Thread 🧶 ⤵️

See details and apply here: careers.roche.com/global/en/…

Join our team as a 12-month PhD intern @ Roche (Basel)! 🧠 Project: Use causal inference & ML on proteomics to find causal drivers of Alzheimer's. For: 3rd-year PhD candidates #Genomics #MachineLearning #CausalInference #CompBio #Alzheimers See details & apply below ⬇️

Pfizer’s COVID data isn’t hidden. It’s been reviewed by FDA, published in peer-reviewed journals, analyzed worldwide, and tracked in millions of people through real-world safety and effectiveness studies. Pretending otherwise is gaslighting, not “gold standard science.”

RFK Jr cut half a billion dollars in research funding for mRNA vaccines. He & Marty Makary claim mRNA vaccines are “risky” so they’ll shift to “traditional, older” vaccine tech that’s “safer.” We used to immunize people by blowing powdered smallpox scabs up their noses. 1/

People don’t wear seatbelts because they prevent all crashes. They wear them because they reduce the risk of death in a crash. COVID vaccines work the same way: they don’t stop every infection, but they dramatically reduce severe disease and death. And they’re not unsafe.

RFK Jr. saying "mitochondria challenge" is a perfect illustration of our time. It sounds sciency. But it's meaningless. It's jargon to create the illusion of knowledge. You see this ALL the time in the pseudo health world. And research shows it works to fool others

The MAHA assault on vaccines will kill people & usher in a scientific Dark Age. RFK Jr has got to go. Join me, my fellow Accountability Journalism institute advisors, Nobel laureates, National Academy members, & colleagues demanding his resignation. change.org/p/demand-rfk-jr-r…

Our didactic review on machine learning for causal inference, now open access We explain • identifiability (theory of when the data can answer a causal question) • machine-learning estimators • study design (asking well-framed questions, and loopholes, eg with time-wise data)

Gorgeous city of Firenze during 12k pre-breakfast run, getting sunlight and downloading good mood before diving into my day 👟💨☀️

Where does Alzheimer’s begin in the brain? Our new study in @NeuroCellPress reveals that earliest changes start deep in cortical layers 5/6, selectively affecting parvalbumin interneurons via reduced NPTX2 & GluA4. Restoring NPTX2 reversed deficits. cell.com/neuron/fulltext/S08…

🚨 Preprint Alert 🔔 Does EEG brain-age work as a neurodegeneration biomarker? Brain age was surprisingly under-predicted in neurological populations. Likely due to diverging age trends in EEG power, breaking model assumptions. Preprint: biorxiv.org/content/10.1101/… #biomarker #EEG

10/ Thanks to all study participants and involved family members, @DPUKDataPortal for providing data access! Big shout out to all co-authors: Sinead Gaubert (lead!), Pilar Garces, Jörg Hipp, Ricardo Bruña, Maria Eugenia Lopéz, Fernando Maestu, Delshad Vaghari, Rik Henson, Claire Paquet

9/ Worried about site effects, controlling for MMSE, spatial averaging, impact of head alignment or comparisons VS healthy controls / cortical slowing? We got you covered by a dozen of supplementary analyses: thelancet.com/cms/10.1016/j.…

8/ Combining all inputs provided better cross-validation scores than any modality by itself (a). We then used conditional permutation importance to inspect the full model (b). MRI freesurfer volumetric measure, MEG covariance, MMSE and site contributed independent information.

7/ We thank multiple unknown reviewers for motivating us to do a more comprehensive analysis, at the risk of using ML on small data. We used a well tested stacking approach (e.g. elifesciences.org/articles/5…) to have a simple multimodal model while not pre-selecting any inputs.