New preprint (and shameless rhetorical question use) in collab with @neilmdavies.bsky.social Commentary on attributable fractions that approach 100%, with lessons for epidemiologists, statistical genetics, etc. More context in thread... #episky zenodo.org/records/2034...

Join us next week for the @DEMONNetworkUK #Genetics & #Omics meeting 🧬 📅 Thursday 14 May ⏰ 2:30 PM (UK time) Delighted to welcome @dylwil.bsky.social as our invited speaker: 👉 “Considering the case for #gene #therapy to address #Alzheimer’s disease” DM for details. #Dementia

Last few places available for our new in-person short course in genetic & molecular epidemiology at University College London, 19th-21st May. Link for registering and more info below... onlinestore.ucl.ac.uk/conferences-...

Our in-person Mendelian randomisation course is back at University College London this autumn! Join us in Bloomsbury, 15th-17th Sept 2026 Tutors: @emmylooroll.bsky.social, me & @neilmdavies.bsky.social Guest lecturer: Eleanor Sanderson, U.Bristol Sign up: onlinestore.ucl.ac.uk/conferences-...

A couple of months ago, I discussed our recent paper on APOE on NPR's Science Friday podcast. Available to hear here or in the usual apps for podcast streaming: www.sciencefriday.com/segments/alz...

Clarifying the scope of gene editing for Alzheimer's disease prevention In response to "Heritable polygenic editing: the next frontier in genomic medicine?" by Visscher et al in Nature last year. zenodo.org/records/1849...

Excited for this one!

Stressing to public the overlapping roles of both environment and genes is fine...it is naive scientific responses like the one pictured (cited by the Mail) we took issue with. Taking away risk broadly will remove disease (at population level). Tho we could consider gene therapy over magic to do so

My fave commentary on our paper in the Mail last week Interested to know what @alucassen.bsky.social uses as a definition of causation though! Here's mine: ajph.aphapublications.org/doi/full/10.... Further reading WRT the topic: zenodo.org/records/1796...

Well done to our colleague @dylwil.bsky.social

Out now!! www.nature.com/articles/s44... Intro to this paper pictured below Huge thanks to collaborators @neilmdavies.bsky.social, @emmylooroll.bsky.social at @uclbrainscience.bsky.social & Sami Heikkinen and Mikko Hiltunen at @uniuef.bsky.social ... #endalz #episky #medsky

Important review by great colleagues in @uclbrainscience.bsky.social

I'm starting a thread on misunderstandings about the role of APOE in Alzheimer's disease / dementia, from the POV of epidemiology / statistical genetics. I will add to this sporadically over time... #endalz #episky #medsky

"Understanding apolipoprotein E as a predominant cause of Alzheimer’s disease" Preprint: zenodo.org/records/1796... Key takeaways below Thx to coauthors @emmylooroll.bsky.social @neilmdavies.bsky.social & N.Mukadam at DOP @ucl.ac.uk Funding @alzheimersresearchuk.org #endalz #episky #medsky

Beware submiting to Scientific Reports... this is what the Research Square page for our article looks like. Yet to get a first decision on it, several months after peer reviews apparently came in and EIGHT after the transfer For-profit publishing racket, which doesn't even bother to publish

@gwascatalog.bsky.social in your study browser, could you plse add a column showing case Ns next to total N for GWAS of diseases? The largest GWAS listed for diseases are now often numerous analyses of biobanks, many with small case Ns. More informative GWAS from case-control consortia get buried

🚨 NEW: 45% of UK adults agree that scrapping the two-child limit is the most effective way to tackle child poverty. When the public get more info about the two-child limit, they are in favour of it being scrapped, which is the action the UK Govt correctly took at the Budget. More: bit.ly/48K2WKD

📢 Registration is open for Causal Inference in Practice — a 3-day short course at UCL 🗓️ 18–20 February 2026 Taught by me, @kdiazordaz.bsky.social & Peter Martin. Learn to estimate causal effects from observational data. 👉 www.ucl.ac.uk/brain-scienc... #CausalInference #Epidemiology #DataScience

Do any pre-print servers for biomedical sciences accept Perspective or Comment articles? Most seem to refuse these types of M/S. I'd like to give some draft articles indexed / searchable presence whilst they get reviewed (& likely rejected) by journals, i.e. not just posting them to Github

Apply to do a PhD with me, Laura Corbin, Gareth Hawkes and George Davey Smith ( @mendelrandom.bsky.social) in Bristol on identifying rare effects on health using new and innovative methods. Amazing team, great educational environment, lovely city. Forward to your undergrads wrapping up!

Last chance to sign up for our MR course next week folks!

Just came across this graphical abstract for a paper. No idea what they were going for with the whale and rainbow, but at least it is very calming

Finally grappled w/ UKB-RAP to update a UK Biobank analysis in past week. Happy to report my sanity is (just about) still intact and I only wasted ~£8 of credit (£5 on a job left running overnight, £3 on other incompetence). Facilitated greatly by ukbrapR by @lcpilling.bsky.social, big thanks Luke!

I have an idea… journals.plos.org/plosbiology/...

Withdraw support for the current government of Israel

BREAKING | Israel depriving Gazans of ‘human dignity’, says UK and 25 other countries in call to end war now

Fabulous work by Phazha Bothongo! 2 years and several near mental breakdowns for us both in the making! 🧠 Causes and consequences of vascular dementia across the life course: Evidence from a UK Biobank phenome-wide and Mendelian randomization study 🧠 www.medrxiv.org/content/10.1...

Are you looking to learn how to conduct and interpret Mendelian randomization studies? You can register now for our in-person short course at UCL. With @emmylooroll.bsky.social and @dylwil.bsky.social 16th-18th September 2025. Details here: www.ucl.ac.uk/brain-scienc...

I love this paper ... so many insights, inc importance of ubiquitous exposures, a clear explanation for flaws of heritability (and analysis of variance more generally), and clarification of differences in "causes of cases & causes of incidence" by Rose, which I always thought obtuse and misleading

This is VERY big genetics & health news, Regeneron (Pharmaceutical company with v strong genetics in medicine arm) is going to buy 23andme. investor.regeneron.com/news-release...

Our in-person Mendelian randomisation course is back at University College London this autumn. Join us in Bloomsbury, 16th-18th September 2025. Tutors: Emma Anderson, Dylan Williams, Neil Davies Guest lecturer: Eleanor Sanderson onlinestore.ucl.ac.uk/conferences-...

Dislike this statement about AD by CARE programme grant invitation from @wellcomeleap.bsky.social and Lisa Mosconi. i) Commonplace misconception that "genetics = unmodifiable"... #episky #medsky

One year later ... a major update to our manuscript showing that most Alzheimer's disease is attributable to the gene APOE. We expanded it to analyses of four studies, now with data on about 460,000 people in total..... www.medrxiv.org/content/10.1... #Alzheimers #Dementia #Epidemiology #EpiSky

"We knew what we were doing. We knew about the indescribable suffering we were causing. We knew about the crimes we were committing, in real time. We knew, and declared for all the world to hear that this was what justice looked like, this was the face of morality."

🚨 New lifetime dementia risk & 2060 US case projections 🧠 This new 2025 study from John Hopkins updates lifetime risk estimates of dementia for the US. Data projects that 42% of adults who are alive at age 55 will develop dementia before they die, assuming current trends continue. 🧵 1/2

Been enjoying the focus on APOE at #AAICAPOE meeting by @alzassociation.bsky.social this week. Here's my poster showing most AD is attributable to APOE variation in four large studies

Delighted to start as a Senior Fellow for Alzheimer's Research UK this month! As are my two chief mascots (help from Bluey here to keep them still). My research will look for ways to combat the bad effects of a protein called apoE, which is the predominant cause of Alzheimer's disease