I was sad to learn that my postdoctoral mentor, David Botstein, died yesterday. I started with David as a postdoc in 1998, and he had a profound effect on both my life and scientific career. He was a giant in the field of genetics, making seminal contributions in both yeast and human genetics. 1/

An important #rstats post here. (Perhaps a better title: Why tidymodels might not be so tidy) I don't eschew this kind of discussion (|> vs %>%, ...) as R Flame Wars. In fact, it's important to continue to hone the language tools we use to express & implement our ideas/models/graphs.

Many living people carry fragments of Neanderthal DNA, remnants of ancient interbreeding events, with uneven distribution across chromosomes. New work by @sarahtishkoff.bsky.social lab suggests patterns are most consistent with Neanderthal contribution to human populations being highly male biased.🧪

Specifically, the idea of source credibility just doesn't apply straightforwardly to LLM interactions. When I get info from another person I can ask: how does the content relate to their expertise? Their lived experience? Their interests and biases? 8/

What's your favourite plot type? Extra points if it's a cool science type plot. I really like Violin plots with swam overlays.

Happy to share our latest paper, the work of @clarajiang.bsky.social who trained brain-based predictors of depression on the UK Biobank (N=7,500 curated cases and matched controls) using AI/deeplearning as well as efficient statistical learning (BLUP). See thread below for a summary of findings

What even is truth? Seems like it should be the purview of philosophers and poets — clearly not people (me) designing snakemake workflows.

I can't be the only one who is made /uncomfortable/ by the way that razor clams burrow into sand, right? The final moments of this video by Kate Crump on the Oregon coastline are sort of the crowning glory of horror. 😬 But how does this even work? Let's talk about bivalve mobility (if we must).

Retraction notice: “the authors acknowledged that their analyses include coding errors. The central error arose from the inverse hyperbolic sine being calculated based on y+1, … left over from an earlier version of the study in which a log (y+1) approach was used instead of the hyperbolic sine.”

@wytamma.bsky.social 's WASM tools have transformed my experience of teaching Python to first year undergraduate biologists this year. Since last year, I've been teaching ~400 undergrads how to code (functions, lists, dictionaries, loops) over (one hour intro lecture +) two 2-hour practicals. 1/n

Check out our new article in #Circulation on an astonishing genetic variant in the ion channel #SCN5A that enables us to learn about #cardiac-arrhythmia and their treatment with sodium channel blockers. @finngen.bsky.social www.ahajournals.org/doi/10.1161/... Congrats to @jwanner.bsky.social

1. Kevin Gross and I have a new paper out today PLOS Biology. We used economic models based around screening games and the market for unpaid labor to highlight a meltdown cycle threatening peer review.

7. We show that this setup leads to a pernicious positive feedback loop. When submissions increase, more peer reviewers are required. Either current reviewers have to spread their effort thinner, or less qualified reviewers must be recruited. Either way, review accuracy drops.

I saw this when it was a preprint? They have a unique version of the estimand-estimator-estimates workflow, will be useful for may biologists, not just ecologists. We need more case studies for ecologists to emulate and for norms to shift so "regression and storytelling" becomes unpublishable.

#rstats Here's a useful guide to creating publication-ready #ggplot figures to journal specifications, which is often quite fiddly. jaquent.github.io/2026/02/crea...

Good advice from the 'bind_cols' documentation: "Where possible prefer using a join to combine multiple data frames. bind_cols() binds the rows in order in which they appear so it is easy to create meaningless results without realising it." #RStats #DefensiveProgramming

yes i wrote a stats textbook shut up shut up shut up. statistics is a discipline in which everybody has impostor syndrome and if you don't then honestly you're probably lying to yourself

#rstats The Australian and New Zealand Journal of Statistics will be running a special issue commemorating the 25th anniversary of the official release of R. (The Univ. of Auckland was the birthplace of R.) They invited various people to contribute articles, including me. 🧵 1/

Last term I tried an experiment: I walked into my Tech and Design Ethics class, admitted that I had *no idea* what to do about ChatGPT - so I would let them figure it out. As in: their first project was to decide and write the ChatGPT policy for the class. Here's what happened:

“In 2025, sponsored research by faculty generated more than $4.8 billion. Compared to that sum, donors contribute a pittance: even the profligate Mr. Bloomberg doesn’t come close. And yet for some reason it’s the donors rather than the faculty who sit on the board and make the big decisions.“

A day in the life of Brooke Wolford, a statistical geneticist at NTNU (Norway). It's proof that science is both technical and deeply human at the same time. 🧬☕️🐕

Really important point here from Elsje. It is crucial to include information about genetic/familial links when considering causes of disorders but lets not hype them as the ultimate solution.

Sometimes, package like modelling packages initially work with tibbles, but then other functions (simple toy-example: `coef()`) do not. Then you start searching *your* code for issues, until you realise it's the tibble that cannot be processed in the *others* code.

American professors, who wants to come to Oxford for a year or two? Get in quick - deadline soon. www.balliol.ox.ac.uk/vacancy-geor...

Tables in publications are often the worst. Even worse is when cells are merged and data is formatted with lines, fonts, and next to each other, as this is only for visualizing data, not for data use. I hate them. If you're using XLSX files as a supplement, give them a useful format!

The 3 hardest things to learn as a scientist: 1. Trust the data.. especially when it’s not what you expected, 2. Trust the data.. allowing it to change your direction, 3. Trust the data.. but not too much: test with new data at every turn.

Yes, they can hallucinate papers that don't exist, discuss results that seem to be imaginary, and can be confusing and inconsistent. But talking to tenured professors may still be helpful

🚨 Proud to share that our paper on the genetic links between type 2 diabetes (T2D) and hypertension is now published in @natcomms.nature.com! We dissect the genetic relationship between T2D and hypertension and show it is not driven by a single shared pathway. 🧬 (1/N)

#Statistics thought of the day: If you think you can find new disease subtypes by empirically clustering patients, think again: www.fharrell.com/post/cluster... #StatsSky

Wow this is a breakthrough in validation of predictive models that fully accounts for point estimates used in predictions being only point estimates. Great work @richarddriley.bsky.social and colleagues! #Statistics #StatsSky

“on average, only about 60% of the information is retained after dichotomization.”

Over the course of 2025, UKB researchers were forced to move all of their data analysis to the cloud rather than working with local copies. What was the effect of this move on scientific output? Here is a simple first pass at answering that question.

New neuromapr #rstats package is available from GitHub! Highly experimental, early adopters and bug identifiers are super welcome to report issues! github.com/lcbc-uio/neu... #neuroscience #rstats #neuroimaging

Very interesting perspective article by @caina89.bsky.social et al on genetic studies in psychiatry. The authors argue that studies increasingly rely on "shallow" phenotyping (self-reported), leading to biases in estimation of genetic relationships between disorders. www.nature.com/articles/s41...

“How do we convince people that learning, itself, is the goal, and to put in real effort when there may be easy ways to game the evaluation system (i.e. grades)?“ #AI #Learning #Education #LLMs

i'm kinda excited about markup languages for cooking: cooklang.org

The program for the 2026 New York Population Genetics meeting, hosted by at the @simonsfoundation.org on March 9th 2026, is now up: events.simonsfoundation.org/event/7c91dd....

Very nice work led @drghawkes.bsky.social, showing convergence between rare and common genetic associations, using whole-genome sequencing data to evaluate the contribution of rare non-coding variants for commonly studied anthropometric traits 🧪🧬 www.nature.com/articles/s41...

Please tell friends & colleagues about our unique course “Genetics & Neurobiology of Language” July 27-Aug 3 2026. Expert tutors, interactive talks, panel discussions, all in a beautiful setting. Scholarships available: meetings.cshl.edu/courses.aspx... @cshlnews.bsky.social @cshlbanbury.bsky.social

Juan Celedon shared this favorite quote during his seminar today: "Never trust anyone's data, particularly your own" #DataScience #QualityControl #Science

Our multi-ancestry GWAS of EXTernalizing conditions (ADHD, substance use...) in ~4M people reveals neurodevelopmental risk, drug-repurposing targets, and yields one of the strongest psychiatric polygenic indices yet! 🧬🎉 doi.org/10.64898/202...

#rstats RIP, John Fox @yihui.org just published this lovely tribute to John Fox and his work yihui.org/en/2026/02/j...

I just whipped up this little #QuartoPub site last week that demonstrates how I teach p-values/hyp-testing through simulation both with live OJS and with #rstats, and I think it's super neat! It has examples for diff-in-means, diff-in-props, and regression slopes nullworlds.andrewheiss.com #statsky

‘Novice workers who rely heavily on AI to complete unfamiliar tasks may compromise their own skill acquisition… We find that AI use impairs conceptual understanding, code reading, and debugging abilities, without delivering significant efficiency gains on average.’ arxiv.org/pdf/2601.20245

Fun news! @gcbias.bsky.social and I are teaching a 2-week online population genetics workshop this summer to raise money for the Center for Population Biology at UC Davis. We're trying to gauge interest -- please fill this out if you might be interested! And please share broadly!

Another Claude project: a static site that pulls in GWAS SNP data from ensemble, multiple public biobanks, open targets, gtex, eqtl catalog, and OMIM. sashagusev.github.io/gwas_lookup/

🧬 New preprint alert! After years of collaborative work across 52 datasets we are presenting eQTLGen phase 2: a genome-wide eQTL meta-analysis covering 43,301 blood samples: www.medrxiv.org/content/10.6... (1/8)

The Exeter team seem to be the Alex Honnold of human genetics. Most sane people on seeing the challenge of meta analysing >1billion variants & 10,000s aggregates across ~100% of the genome would opt for the safety rope of the exome. Not these guys.

Footnote: "The em dashes reflect our stylistic preference and should not be interpreted as evidence of Al-assisted text generation."

We're excited to be recruiting an NIH funded postdoc to work in the Coop lab at UC Davis. We're specifically interested in candidates who are want to work at the intersection of human genetics, GWAS, and population genetics modeling. Please RT