Sign in

Kasper D Hansen

@kasperdhansen.bsky.social
2.1K followers 118 following 64 posts

Professor @ JHU, statistical genomics, www.hansenlab.org

PostsRepliesMedia
Reposted by Kasper D Hansen
Johns Hopkins Dept of Biostatistics @jhubiostat.bsky.social · 22/09/2026
Come work with the Johns Hopkins Dept. of Biostatistics! We are currently hiring two open positions: Postdoctoral Fellow (Data Science and AI in Health & Medicine) and Assistant Professor (tenure track). Learn more online: publichealth.jhu.edu/departments/...
We Are HIRING
Join Our Department

Open Positions:
Assistant Professor
Postdoc: Data Science and AI in Health & Medicine
032
Kasper D Hansen @kasperdhansen.bsky.social · 10/09/2026
So humans cannot produce "slop"? Seems wrong to me. I agree there is a difference between "slop" and "crap" though.
111
Reposted by Kasper D Hansen
American Society of Human Genetics (ASHG) @geneticssociety.bsky.social · 04/09/2026
@ajhgnews.bsky.social sat with Kimberley Anderson, PhD, in the latest "Inside AJHG" to discuss her recently published paper, “Androgens mediate sexual dimorphism in Pilarowski-Bjornsson syndrome.” ➡️ bit.ly/4hbuxsp #ASHG #HumanGenetics #GeneticsDiscoveries
044
Reposted by Kasper D Hansen
Needhi Bhalla 💅🏽 @needhibhalla.bsky.social · 03/09/2026
absolute fuckers "Trump officials at NIH disclosed the existence of DoD’s efforts only after being pressed for information by House Appropriations Committee Democrats." democrats-appropriations.house.gov/news/press-r...
democrats-appropriations.house.gov
Defense Officials Attempt to Raid NIH Funds to Cover Department of Defense Activities
House Appropriations Committee Ranking Member Rosa DeLauro (D-CT-03) has learned that officials at the Department of Defense (DoD) are working to divert billions of dollars from the National Institute...
1018394
Reposted by Kasper D Hansen
The American Journal of Human Genetics @ajhgnews.bsky.social · 03/09/2026
🧬New from Anderson et al! 📄Androgens mediate sexual dimorphism in Pilarowski-Bjornsson syndrome
cell.com
Androgens mediate sexual dimorphism in Pilarowski-Bjornsson syndrome
Pilarowski-Björnsson syndrome (PILBOS) is a neurodevelopmental disorder caused by pathogenic variants in CHD1. Here, we characterize a large PILBOS cohort and describe their variants. Using a patient-specific mouse model, we demonstrate that androgens protect against CHD1 deficiency, suggesting a mechanism for sex-specific penetrance in an autosomal disorder.
011
Kasper D Hansen @kasperdhansen.bsky.social · 24/08/2026
The new boquerones from Trader Joe are excellent.
001
Reposted by Kasper D Hansen
Jenna Norton @jenna-m-norton.bsky.social · 20/08/2026
“Flailing regimes dehumanize minorities to distract from their failures.” - Julia Curlee www.theatlantic.com/magazine/202...
theatlantic.com
The Wrong Kind of American
The Trump White House needed me—until the fact that I am trans became public.
02911
Reposted by Kasper D Hansen
sebmal @mrdo.pro · 28/07/2026
WARNING! TACO BELL ADVISES NEW TACO SO GOOD IT IS BEYOND THEIR CONTROL, COULD REPLACE ALL OTHER FOOD BEFORE HUMANS CAN CONTAIN IT. URGES CONGRESS TO GIVE THEM MORE MONEY SO THEY CAN DEVELOP ANTI-TACO SAFEGUARDS
12796214
Reposted by Kasper D Hansen
Leandros Boukas @leandrosboukas.bsky.social · 16/07/2026
Together with Hans Bjornsson (and @kasperdhansen.bsky.social, and others) we have spent many years thinking about the molecular basis of Mendelian disorders caused by variants in chromatin regulators. We have written a review where we discuss our thoughts: www.tandfonline.com/doi/full/10....
tandfonline.com
Mendelian disorders of the epigenetic machinery: a decade of insights into the molecular basis
The Mendelian disorders of the Epigenetic Machinery (MDEMs), or Chromatinopathies, are now understood to be a collectively common cause of childhood neurodevelopmental delays and intellectual disab...
063
Kasper D Hansen @kasperdhansen.bsky.social · 09/07/2026
Liz Heller at UPenn is great on the molecular mechanisms.
020
Reposted by Kasper D Hansen
Michael Hoffman @michaelhoffman.bsky.social · 11/06/2026
We need to think bigger for our Project 2029. Current Democratic efforts don't rise to the challenge. Worth reading this column by @jamellebouie.net even moreso than usual.
091
Reposted by Kasper D Hansen
Michael Okun @michael-okun.bsky.social · 10/06/2026
I’ve officially resigned as Associate Editor for Frontiers in Systems Neuroscience. It used to be a reputable journal, but became a case study in how forced automation destroys academic integrity. 👇
5435861430
Reposted by Kasper D Hansen
Carl T. Bergstrom @carlbergstrom.com · 04/06/2026
11. Addendum: here are the words characteristic of LLMs. Notice anything? It reads like an email from our upper administration.
Table S1. List of 228 Focal Words disproportionately associated with LLM outputs.3
4
achieves contextualize exacerbates integrates oversight showcasing
achieving contextually exacerbating integrating oversimplify signifying
acknowledges cornerstone exceeding intensifies paving situates
adaptable crucial excelled intensifying pinpointing situating
addressing culminating excelling interpretability pivotal sourced
adept delineates excels interpretable poised strategically
adeptly delve exemplifies intricacies posits streamline
adjustments delved exhibits intricacy practicality streamlined
adjusts delves facilitating intricate preserving streamlining
advancements delving featuring intricately primarily strengthens
advancing dependencies formidable introduces prioritise struggle
advocating diminishes fortifying juncture prioritize stylistic
aiding discernible fostering leverage prioritizes surpasses
align downturns fosters leverages prioritizing surpassing
aligning dynamically foundational leveraging promise symbolizing
aligns elucidates frameworks meticulous prowess synthesizes
alongside elucidating garnered meticulously reaffirming thereby
amidst emphasising garnering mitigate realm thoroughness
amplifies emphasizes grappling mitigates redefines thoughtfully
avenues emphasizing groundbreaking mitigating refine thrives
boasting employs harnessing multifaceted refines timelines
boasts enabling heighten navigates refining transcends
bolster encapsulates heightened navigating reflecting transitioning
bolstering encompass heightening necessitate reinforces underexplored
bolsters encompassed heightens necessitates reliance undermines
broader encompasses highlighting necessitating renowned underperform
capitalizes encompassing highlights notable reshapes underperformed
captivating endeavors hinges notably reshaping underpins
captures enhance imperatives nuanced revolutionize underscore
capturing enhancements inadequately nuances revolutionizing underscored
categorizes enhance…
2126165
Kasper D Hansen @kasperdhansen.bsky.social · 03/06/2026
Absolutely phenomenal infrastructure work by Dirk, one of the giants of the R community. This kind of work is of high importance, but (usually) does not get the recognition it deserves. Thank you for your continued commitment to excellence.
1111
Kasper D Hansen @kasperdhansen.bsky.social · 28/05/2026
As I read it, there are some big changes sneaking into this notice.
023
Kasper D Hansen @kasperdhansen.bsky.social · 26/05/2026
Finally, I am not really familiar with the primary literature in plants. I am happy to be pointed to the key work that we should have cited.
000
Kasper D Hansen @kasperdhansen.bsky.social · 26/05/2026
In my view, this note really belongs in the paper, but we had to cut >60% of our text following acceptance.
100
Kasper D Hansen @kasperdhansen.bsky.social · 26/05/2026
We are citing two review articles that discuss paramutations in plants. There is a better discussion of previous work in the Supplementary Note "On various intergenerational epigenetic inheritance patterns"; still no citation of the primary literature in plants.
100
Kasper D Hansen @kasperdhansen.bsky.social · 26/05/2026
I have been thinking about whether I should reply to this post, since I don’t have much to say. But in the end, I want to acknowledge the criticism. Anyone can check our references and confirm that @maxvcg.bsky.social is right: there is no citations to the primary literature in plants.
200
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
I forgot to say the for the (newer) basecaller we use for muscle, there is no use of a reference genome, so the concern essentially goes away.
010
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
Now, for sure, the reference genomes we use are far from perfect. One could absolutely be concerned about areas where the reference genomes are less good. Here, I return to my initial observation that this really affects cis-patterns.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
Additionally, we are focused on regions. That doesn’t mean that we’re immune to such an effect, but inside each DMR we have many CpGs, most of which would be >5bp away from any single variant as described by deCODE. In my view, this is an inherent advantage of a region-based approach.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
For muscle, we use a completely different chemistry and basecaller and there has been a substantial improvement in DNA methylation calling because of improvements in both chemsitry and computation, so it is not clear to the extent that this concern is still valid.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
This is therefore a vastly different situation from the deCODE paper where they study humans with a single reference genome.
110
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
We phase the data first, and we then use Nanopolish with the appropriate strain specific genome. This ensures there is no variation between the sample genome and the reference genome used by Nanopolish except where the strain reference genome is not correct (see below for this case).
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
For liver, we used R9 chemistry and Nanopolish to get DNA methylation. This is similar to the deCODE paper. In Nanopolish, the data is first base called into ACGT nucleotides and the, subsequently, Nanopolish adds DNA methylation information using the basecalled reads and a reference genome.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
Second, what did we actually do? This is going to be a long answer because we used 2 different chemistries and 2 different basecallers for the 2 different tissues. IMO this is a strength of our study because it provides a natural layer of robustness for the computational pipeline.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
Many of the more interesting patterns are therefore provably not affected by this concern, because they are not driven (mostly or entirely) by local sequence difference.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
First, a systematic error driven by local genotype would look like a cis-meQTL, where the methylation is driven entirely by local sequence. This concern is therefore mostly about a big class of - I would say very well established - patterns which we spend relatively little energy on.
100
Kasper D Hansen @kasperdhansen.bsky.social · 20/05/2026
The (best) brief explanation is that there is a big difference between sequencing humans compared to sequencing inbred mice where you have a strain-specific genome. In what follows I'll provide a lengthy discussion of this potential issue and why I am personally not too worried.
110
Reposted by Kasper D Hansen
Robin Andersson @randersson.bsky.social · 19/05/2026
Excited to see this preprint out from the lab! Congrats to @heinarsson.bsky.social, Natsuda, @robertkrautz.bsky.social, and all co-authors involved in this major effort on developing nucCAGE and PRIME, and on evaluating transcription initiation-defined regulatory elements for variant interpretation!
biorxiv.org
0186
Kasper D Hansen @kasperdhansen.bsky.social · 14/05/2026
Yeah I don’t see why this is a valid cause for rejection.
030
Kasper D Hansen @kasperdhansen.bsky.social · 11/05/2026
Really interesting critique of a paper in Science. By now it is well established that you need to compare to simple baselines before making big claims.
i.science
060
Reposted by Kasper D Hansen
Nicholas Mancuso @nmancuso.bsky.social · 07/05/2026
1263
Reposted by Kasper D Hansen
Bioconductor @bioconductor.bsky.social · 30/04/2026
🚀 We’re excited to share the release of Bioconductor 3.23, featuring: ✅2418 software packages (94 new) ✅437 experiment data packages (5 new) ✅928 annotation packages ✅8 books (2 new) ✅28 workflows 🔗 Please see the full release announcement here: bioconductor.org/news/bioc_3_...
02413
Reposted by Kasper D Hansen
Darren Dahly @statsepi.bsky.social · 23/04/2026
JPAI expresses concern that AI will generate even more papers than he does. 😜
2205
Kasper D Hansen @kasperdhansen.bsky.social · 14/04/2026
In my (perhaps limited) experience, monitors cannot be reduced just to stats. I say go with what you know you like, despite the price.
010
Kasper D Hansen @kasperdhansen.bsky.social · 21/03/2026
I hope you prevail.
010
Reposted by Kasper D Hansen
Benoit Bruneau @benoitbruneau.bsky.social · 21/03/2026
this truly excellent. Having lost Tanya 3.5 years ago, my father a few weeks ago, and my dog last week (not kidding), this is timely
9442
Kasper D Hansen @kasperdhansen.bsky.social · 06/03/2026
This is really impressive.
020
Reposted by Kasper D Hansen
Jenna Norton @jenna-m-norton.bsky.social · 06/03/2026
Current NIH leadership want you to think they are using rigorous, consistent & scientific processes to screen studies to align them with agency priorities. But the process that they have put down on paper is a sham. It’s important to know NIH is not following its own guidance. Here’s why: 🧵1/
4224131
Reposted by Kasper D Hansen
The American Journal of Human Genetics @ajhgnews.bsky.social · 05/03/2026
🧬 New from @leandrosboukas.bsky.social & co 📄Identification of de novo variants from parent-proband duos via long-read sequencing 🖥️ github.com/sbergercnmc/...
cell.com
Identification of de novo variants from parent-proband duos via long-read sequencing
Determining de novo status is often what upgrades a variant from VUS to pathogenic, but it has not been possible without sequencing both biological parents. We developed, extensively evaluated, and ap...
042
Reposted by Kasper D Hansen
Michelle Cohen @docmcohen.medsky.social · 28/02/2026
The ultimate goal of wellness consumerism is to shift all responsibility to the individual health consumer who then gets the "freedom" of navigating an unregulated snake oil market.
02611
Kasper D Hansen @kasperdhansen.bsky.social · 04/02/2026
How about porting Linux to Rust?
110
Reposted by Kasper D Hansen
Bruce Springsteen @brucespringsteen.net · 28/01/2026
I wrote this song on Saturday, recorded it yesterday and released it to you today in response to the state terror being visited on the city of Minneapolis. It’s dedicated to the people of Minneapolis, our innocent immigrant neighbors and in memory of Alex Pretti and Renee Good. Stay free
youtu.be
Bruce Springsteen - Streets Of Minneapolis (Official Audio)
YouTube video by Bruce Springsteen
42569708940549
Kasper D Hansen @kasperdhansen.bsky.social · 06/01/2026
I have not tested it. I strongly think action=error should be default.
000
Reposted by Kasper D Hansen
Hongkai Ji @jihk.bsky.social · 19/12/2025
Glad to share our benchmark study in Genome Biology on highly variable gene (HVG) selection methods for single-cell RNA-seq. We show that a hybrid strategy performs best, now implemented in the mixhvg R package. rdcu.be/eUfjt
rdcu.be
A systematic evaluation of highly variable gene selection methods for single-cell RNA-sequencing
051
Reposted by Kasper D Hansen
Jake Berg @jacobsberg.bsky.social · 18/12/2025
Gift link, probably the funniest thing the Wall Street Journal has done this year.
wsj.com
We Let AI Run Our Office Vending Machine. It Lost Hundreds of Dollars.
An AI agent ran a snack operation in the WSJ newsroom. It gave away a free PlayStation, ordered a live fish—and taught us lessons about the future of AI.
9730061044
Reposted by Kasper D Hansen
James McInerney @jomcinerney.bsky.social · 10/12/2025
This is shocking stuff. John Quackenbush has been brilliant since forever. www.statnews.com/2025/12/05/r...
statnews.com
His lab was humming with discovery. After one year under Trump, it’s almost silent
John Quackenbush built a lab that is at the forefront of human genetics research and bioinformatics. Trump administration cuts have put it in danger of collapse.
0109
Reposted by Kasper D Hansen
Brendan Nyhan @brendannyhan.bsky.social · 20/11/2025
This is real x.com/grok/status/.... Yangtze record swim times but in LLM form
85311