Gerome Breen

GLAD is now >30,000 volunteers. Thanks to all of our participants, supporters, and all the researchers who have helped with recruitment in >200 clinical sites! If you have ever experienced depression or anxiety. @GLADStudy @NIHRBioResource @NIHRMaudsleyBRC

Ireland’s population chart remains a wild one to look at. The country has still not recovered from the Great Famine (1845-52).

New @KingsCollegeLon @NIHRMaudsleyBRC research has split #schizophrenia risk into 2 genetically distinct pathways: ▪️1 shared with #BipolarDisorder & linked to better educational outcomes ▪️1 unique to schizophrenia & linked to poorer educational outcomes kcl.ac.uk/news/advanced-…

Polygenic risk scores for heart disease, which incorporate hundreds of common genome variants, are capable of identifying risk for patients well beyond traditional risk factors, and benefit of getting ahead of the disease, i.e. prevention. A new review @NEJM . nejm.org/doi/full/10.10… @pnatarajanmd and colleagues

📢@KingsCollegeLon will co-lead 2 of the 4 national workstreams in the £50m government–funded Mental Health Goals programme, helping to transform #MentalHealth data, #ClinicalTrials and #industry partnerships in the UK. 🔗Read more: kcl.ac.uk/news/kings-col…

NYT has an article on the usual gang of quantitative racists misleading the NIH to access genetic data from children. My view is that data should be made as available as possible within the bounds of patient consent, but that lying to access data is egregious academic misconduct.

Yes, London has the lowest murder rate in decades, knife crime has fallen sharply, victims of personal crime are at the lowest level since 2002, but who needs facts and reality when your supporters are gullible morons? x.com/BestForBritain…

@itvpeston @KemiBadenoch @Keir_Starmer @EdwardJDavey Demanding details now a bit amateur - why would they give details when Russia would try to counter them? Good he committed to a vote. Edward Davey - what else could Starmer do given UK's constraints because of BREXIT? Macron a bit stronger but not much: france24.com/en/france/2026…

Today at #PMQs: 🔵 @KemiBadenoch described the lack of detail on plans to deploy British troops to maintain any peace deal in Ukraine as “astonishing” 🌹 @Keir_Starmer committed to a vote in the House of Commons if any deployment were to take place 🔶 @EdwardJDavey said the PM looked “ridiculous” for not accepting President Trump had broken international law in Venezuela What did you make of it? #ThreeWordReviews please 👇 #Peston

The A90 south of Aberdeen looks like something from a video you see of Arctic Norway or Finland. One of the biggest snow events of the last 20 years in north-east Scotland for sure.

Apply for our @KingsIoPPN @KingsCollegeLon postdoc post on the @GLADStudy, the largest genetics study of severe MDD & anxiety (>50K). You’ll do analysis at scale - GWAS/WGS, multi-omics, EHR & voice phenotyping - at @SGDPCentreKCL Apply by 11 Jan 2026 jobs.ac.uk/job/DPW445/gla…

This paper from Harvard and MIT quietly answers the most important AI question nobody benchmarks properly: Can LLMs actually discover science, or are they just good at talking about it? The paper is called “Evaluating Large Language Models in Scientific Discovery”, and instead of asking models trivia questions, it tests something much harder: Can models form hypotheses, design experiments, interpret results, and update beliefs like real scientists? Here’s what the authors did differently 👇 • They evaluate LLMs across the full discovery loop hypothesis → experiment → observation → revision • Tasks span biology, chemistry, and physics, not toy puzzles • Models must work with incomplete data, noisy results, and false leads • Success is measured by scientific progress, not fluency or confidence What they found is sobering. LLMs are decent at suggesting hypotheses, but brittle at everything that follows. ✓ They overfit to surface patterns ✓ They struggle to abandon bad hypotheses even when evidence contradicts them ✓ They confuse correlation for causation ✓ They hallucinate explanations when experiments fail ✓ They optimize for plausibility, not truth Most striking result: `High benchmark scores do not correlate with scientific discovery ability.` Some top models that dominate standard reasoning tests completely fail when forced to run iterative experiments and update theories. Why this matters: Real science is not one-shot reasoning. It’s feedback, failure, revision, and restraint. LLMs today: • Talk like scientists • Write like scientists • But don’t think like scientists yet The paper’s core takeaway: Scientific intelligence is not language intelligence. It requires memory, hypothesis tracking, causal reasoning, and the ability to say “I was wrong.” Until models can reliably do that, claims about “AI scientists” are mostly premature. This paper doesn’t hype AI. It defines the gap we still need to close. And that’s exactly why it’s important.

“It’s so good… It’s wisdom.” Reflecting on his guest edit and leaving In Our Time after almost 30 years, Melvyn Bragg gives an emotional reading of Thomas Hardy's 'She, To Him'.

This is a map of black soil — the most valuable agricultural land in Europe. As you can see, most of it lies in Ukraine. Because of this, Ukraine’s territory was one of the most densely populated areas in the pre-industrial era. At a time when food couldn’t be transported across half the globe, people could survive and flourish only where it grew well. Six thousand years ago, there were settlements of up to 10,000 people on Ukrainian lands. These communities were farmers; they produced linen clothing and developed pottery-kiln technologies unmatched anywhere else in the world for the next two millennia. They left no written language, but they created practical models out of clay — miniature houses showing tools, people, and daily life. And remember: this was the era when the first cities were emerging in Mesopotamia, long before the pyramids were built. Herodotus writes that a Scythian king once asked each subject to bring an arrow so he could count the population — and received seven million. Whether that number is precise or not, it tells us that the region was heavily populated. Alexander the Great sent his general Zopyrion to conquer Scythia; the entire army vanished, along with Zopyrion himself. Alexander was wise enough not to try again — perhaps that’s why he never lost another battle. Ukraine never experienced natural famines. That’s why Ukrainians are so individualistic: if you can grow your own food, you don’t need a “strong leader” to survive. In russia, it was different — harsh, swampy lands forced communities to share scarce food and follow a leader as if he were a god. But we did experience famines — all artificially created by Moscow. In 1933, millions of Ukrainians were deliberately starved when their food was seized. Every Ukrainian family carries a horror story from that time. Today, Ukraine can feed the entire EU and already feeds millions in North Africa and Asia. If russia succeeds in taking over, it could trigger a global crisis simply by cutting off these supplies. Hungry and desperate people will be forced to flee, with Europe becoming their last hope. Are you ready for that? If not — help Ukraine. Author: Volodymyr Kukharenko

😡 Brexiteer Tim Martin spent years telling Britain that Brexit would “boost democracy and prosperity”. Today? Wetherspoons is routing its entire EU expansion through Ireland to avoid the Brexit red tape *HE* helped create. The loudest Brexiteer in hospitality now needs the EU Single Market to grow his business 👉because UK firms outside it can’t compete. Brexit didn’t take back control. It pushed British business to take flight and now Britain 🇬🇧 is rapidly becoming bankrupt and destitute. 👉 independent.ie/business/wethe…

Extraordinary words by Sir John Major on #Brexit at LSE last night: 🧵 🇬🇧 🇪🇺 1. “In an act of collective folly, the United Kingdom voted to lead to European Union across the world, our enemies celebrated and our friends despair.” “We left Europe on a minority vote of 37% of the election after a referendum campaign that was packed with misinformation and misjudgment, it left our country poorer, weaker and divorced from the richest free trade market that history has ever seen. National interest was brushed aside by false hopes and promises”. “False hopes and promises that even a cabinet dominated by front line Brexit enthusiasts was unable to deliver the promises they made, […] while the forecast damage of leaving the European Union has become only too apparent. The nation saw Project Fear become Project Reality very easily. It's no consolation that the majority of the public now overwhelmingly recognizes that it was mislead in their moments of triumph.” “Brexiteers predicted other countries would follow their lead and leave the European Union. None have. All saw only too clearly that Brexit was packed with disadvantages, as we meet far from others leaving the European Union, as we meet nine further nations now wish to join the European Union, which is an apt comment on how the world saw Britain's decision.” “United Kingdom once reveled, not long ago, within the memory of everyone present in this room, once reveled in being a leading member of the European Union with half a billion citizens and the undoubted first ally of the United States, the world's most eminent superpower. Today, we know we are neither, and so does the world”.

🏅 Congratulations George Davey Smith, our Scientific Director, on winning @The_MRC Millennium Medal! 🧬 For exceptional contributions to improving human health through world-class medical research 👉 Read more: bristolbrc.nihr.ac.uk/medical-resear…

Prader-Willi syndrome generally results from loss-of-function of paternally-expressed 15q11-13 while the maternal copy is imprinted, or epigenetically silenced. CRISPR-based epigenomic editing of iPSCs from PWS pts restores expression of the locus from the maternal copy nature.com/articles/s4146… @NatureComms

@timothycbates Yes - I was rereading it recently and realised it is the near future in many ways.

We are *so* close now to the Diamond Age and “The Young Lady’s Illustrated Primer” envisioned by Neal Stephenson! Bring on the Propaedeutic Enchiridion!

Despite Ireland being the EU's biggest per capita exporter of goods to the US, its effective tariff rate (ETR) is the second lowest among the member states. That is because of the composition of exports, dominated by pharmaceuticals which have not been tariffed by the US. From today's Autumn forecast report by the European Commission.

My ten cents worth on ten years of fuckery with Robbie Gibb from my time as editor of @Channel4News to trying to get @BBCNews to broadcast our film on Gaza, through No 10, Brexit and Israel the self appointed defender of impartiality, who has utterly destroyed it.

Preprint from Joshua Au Yeung @richdobson and Zeljko Kraljevic @KingsIoPPN on AI's worrying potential for adverse mental health outcomes. The Psychogenic Machine: Simulating AI Psychosis, Delusion Reinforcement and Harm Enablement in Large Language Models. arxiv.org/abs/2509.10970