Sign in

junhyongkim.bsky.social

@junhyongkim.bsky.social
347 followers 300 following 93 posts

A genomics cat, sometimes Christopher H. Browne Professor of Biology at Penn. All posts my opinion only.

PostsRepliesMedia
Reposted by @junhyongkim.bsky.social
bioRxiv Genetics @biorxiv-genetic.bsky.social · 06/10/2026
Linkage disequilibrium moderates how evolving genetic background shapes variant-level architecture www.biorxiv.org/content/10.64898/20…
092
junhyongkim.bsky.social @junhyongkim.bsky.social · 06/10/2026
What? Nobody asked how 100 trillion cells work together to do all those things from a single fertilized egg?
010
junhyongkim.bsky.social @junhyongkim.bsky.social · 03/10/2026
Public forum for communication is a precious resource. All communication channels are being corrupted by tech (e.g. spam calls). Slowing things down is a good heuristic to create a cost incentive for low quality messages.
1244
junhyongkim.bsky.social @junhyongkim.bsky.social · 03/10/2026
Second, unlike experimental sciences, indeed you can generate 10 meh ideas a day and then run a computational production line in CS/Data Science. Should all this compete in the free market of ideas? Sure go ahead and post it on your website. Tweet all you want.
1120
junhyongkim.bsky.social @junhyongkim.bsky.social · 03/10/2026
There's both a misunderstanding of the problem that arXiv is addressing and the problem generated by the change in policy. First, there is no scientist who is sitting on 3 finished manuscripts, not having previously submitted at least one of them, and suddenly is stymied by the 2/month limit.
1203
Reposted by @junhyongkim.bsky.social
Maitreya Dunham @maitreya.bsky.social · 24/09/2026
Exciting news: @uwgenome.bsky.social will be hiring for two tenure-track faculty positions this year, one open rank and one assistant professor. Both are “Genome Sciences” interpreted broadly so we welcome applicants working in a variety of fields and methods. Ad will be out shortly.
1107105
Reposted by @junhyongkim.bsky.social
Xin Sun @xinsun-putiger.bsky.social · 24/09/2026
I’m deeply honored to receive the 2026 AGU Ocean Sciences Early Career Award! Thank you to AGU, the Ocean Sciences community, and my nominators for this recognition! [1/2]
393
Reposted by @junhyongkim.bsky.social
Steve Joffe @stevejoffe.bsky.social · 14/09/2026
Thrilled to share that Penn Medical Ethics @pennmehp.bsky.social is inviting applications for our 3-year National Human Genome Research Institute-funded postdoctoral fellowship in the ethical, legal, & social implications (ELSI) of genomics. Info @ medicalethicshealthpolicy.med.upenn.edu/elsipostdoc
medicalethicshealthpolicy.med.upenn.edu
Penn Medical Ethics and Health Policy | ELSI Genomics Postdoctoral Training Program
About the Fellowship The mission of the 3-year Penn Postdoctoral Training Program in the Ethical, Legal and Social Implications (ELSI) of Genetics and Genomics, supported by a training grant from...
143
junhyongkim.bsky.social @junhyongkim.bsky.social · 12/07/2026
One of the most profound articles I've read. "Just try to be decent"
000
Reposted by @junhyongkim.bsky.social
Matthew Taliaferro @jmtali.bsky.social · 11/06/2026
New paper! How do RNAs "know" where to go inside a cell? We dug into the sequence elements that route RNAs to the right place. It turns out that, in mammals, they're surprisingly massive (>200 nt), multipartite, and wonderfully complicated. 🧵
816565
Reposted by @junhyongkim.bsky.social
Christine Mayr @christinemayr.bsky.social · 08/06/2026
Finally out in @Cellcellpress! Proteins with long IDRs are prone to misfolding during protein synthesis. This is prevented by mRNA 3′UTRs that act as mRNA-based IDR chaperones. www.cell.com/cell/fulltex...
cell.com
mRNA 3′ UTRs chaperone intrinsically disordered regions to control protein activity
Highly conserved mRNA 3′ UTRs act as co-translational chaperones for intrinsically disordered regions (IDRs), preventing inter-domain misfolding and enabling biogenesis of fully active proteins.
522597
Reposted by @junhyongkim.bsky.social
Caroline Bartman @cbartman.bsky.social · 20/05/2026
🚨first Bartman lab preprint! it's an LC-MS method to measure "total Coenzyme A" (sum of all acyl-CoAs). this will let us study coA synthesis flux using heavy-isotope tracing! huge congrats to first author Amelia Taylor and star coauthor @mzspectrum.bsky.social ! www.biorxiv.org/content/10.6...
biorxiv.org
1014827
Reposted by @junhyongkim.bsky.social
Mark Peifer (He, him) @peiferlabunc.bsky.social · 06/04/2026
If you know a graduating undergraduate looking for the chance to spend 1-2 years in a lab before graduate school, we're hiring! Our lab explores morphogenesis, defining how the cell adhesion & cytoskeletal machinery work together to allow cells to change shape & move tarheels.live/peiferlab/ 1/n RT
312887
Reposted by @junhyongkim.bsky.social
S Otto @sarperotto.bsky.social · 31/03/2026
Insightful summary of an extraordinary person being extraordinarily fired: "you don't silence a woman of color who raised concerns about racism by silencing her...The fingerprints are distributed perfectly across the system." www.linkedin.com/pulse/uclas-...
linkedin.com
UCLA's Most Decorated Ecologist Just Got Fired : Here's the sequence
In January of 2025, at the ICTP ICTS winter school at NCBS, Priyanga Amarasekare offered to take me and a few other students out for my birthday. She didn't have to do that.
14337
junhyongkim.bsky.social @junhyongkim.bsky.social · 28/03/2026
When the tech companies built their entire empire on algorithmically pushing what people want to hear to confirm their biases, why would they not do the same with LLMs
000
Reposted by @junhyongkim.bsky.social
Rita Strack @ritastrack.bsky.social · 25/03/2026
Well here's an answer to a question I hadn't thought of asking www.nature.com/articles/s41...
nature.com
Limitations of serial cloning in mammals - Nature Communications
Here they show that extended serial somatic cell cloning imposes a threefold increase in de novo mutations compared to natural reproduction, progressively reducing birth rates and ultimately limiting ...
2286
junhyongkim.bsky.social @junhyongkim.bsky.social · 21/03/2026
Wow literally 100+ cool places to go dependent on your likes. I always take friends to this bar: www.chunnyun.com
chunnyun.com
020
Reposted by @junhyongkim.bsky.social
Mir Lab @mirlab.bsky.social · 20/03/2026
@mukherja.bsky.social work on RNAPII clusters is in print today: www.nature.com/articles/s41... Congrats! TLDR: RNAPII clusters represent transcriptonally engaged molecules at single genes, not super-stochiometric assemblies. Please check it out and see the the thread below for a summary.
nature.com
A single cluster of RNA Polymerase II molecules is stably associated with active genes - Nature Communications
Live molecular-scale imaging in fly embryos shows that a single RNA Polymerase II cluster stays bound to active genes during transcription. Cluster formation requires initiation, and intensity represe...
04917
Reposted by @junhyongkim.bsky.social
Colin Conine @colinconine.bsky.social · 19/03/2026
There is accumulating evidence that sperm miRNAs function to regulate embryonic gene expression and transmit inherited phenotypes to offspring but how does this work since sperm cargo is massively diluted in the egg upon fertilization. Here we directly address this ? www.biorxiv.org/content/10.6...
biorxiv.org
33212
Reposted by @junhyongkim.bsky.social
Tugce Aktas @aktast.bsky.social · 25/02/2026
Our most recent work on the “function and evolution” of #nuclear-speckles is now online at Cell @cp-cell.bsky.social doi.org/10.1016/j.ce... Read the thread👇 for the highlights of our findings.
doi.org
Redirecting
912561
junhyongkim.bsky.social @junhyongkim.bsky.social · 26/02/2026
Unfortunately, whenever I think about posting something science and login and see all that's going on, it just doesn't seem that important. Hard not to feel that way about doing science too...
011
junhyongkim.bsky.social @junhyongkim.bsky.social · 23/02/2026
Free markets are supposed to select for the good businesses and kill the bad. But, the hyper efficient capital markets allow transient feasting & abandoning the dying carcass. What we have now is a bunch of parasites on the economy evolving unbridled virulence. Now the viruses want to rule.
141
Reposted by @junhyongkim.bsky.social
Yun S. Song @yun-s-song.bsky.social · 21/02/2026
Can we simulate realistic evolutionary trajectories and “replay the tape of life”? In this work, we propose a flexible, generalizable deep learning framework for modeling how the entire protein sequence evolves over time while capturing complex interactions across sites. 1/n doi.org/10.64898/202...
doi.org
38735
Reposted by @junhyongkim.bsky.social
Seychelle Vos @voslab.org · 19/02/2026
🚨New preprint! We built a cell-free genomics platform (GATO-seq) to probe transcriptional regulation and discovered a “super pause” sequence that triggers a new Pol II active-site conformation. Huge shout-out to @robertovn.bsky.social for pulling off this monster of a project. tinyurl.com/superpause
04912
junhyongkim.bsky.social @junhyongkim.bsky.social · 16/02/2026
The problem is, hiring based on feels can lead to shutting out those who don't fit the expected cultural phenotype. Its hard to thread that needle between exercising bias and warding off the creeps.
010
Reposted by @junhyongkim.bsky.social
Lukasz Bugaj @bugajlab.bsky.social · 02/02/2026
Now in print at ACS Synbio! Introducing Aviatar: a concept for inducible translocation to any cell compartment using only a 1-component probe pubs.acs.org/doi/10.1021/...
pubs.acs.org
Optogenetic Translocation to Subcellular Compartments through Regulation of Protein Avidity
Inducible translocation to subcellular compartments is a common strategy for protein switches that control a variety of cell behaviors. However, existing switches achieve translocation through induced...
0196
Reposted by @junhyongkim.bsky.social
The Warren Center for Network & Data Sciences @warrencenter.bsky.social · 27/01/2026
Warren Center faculty affiliate Greg Ridgeway has developed and tested a statistical model using real-world data that could help police departments take steps to prevent future excessive force. penntoday.upenn.edu/...
011
junhyongkim.bsky.social @junhyongkim.bsky.social · 27/01/2026
Ha ha. Almost reads like a Haiku
010
Reposted by @junhyongkim.bsky.social
Rita Strack @ritastrack.bsky.social · 23/01/2026
Well that's cool Source: Nature search.app/FLn5P
search.app
Quantum spin resonance in engineered proteins for multimodal sensing - Nature
A recently developed class of magneto-sensitive fluorescent proteins are engineered to alter the properties of their response to magnetic fields and radio frequencies, enabling multimodal sensing of b...
2283
Reposted by @junhyongkim.bsky.social
Carolyn Bertozzi @carolynbertozzi.bskyverified.social · 19/01/2026
This is how we first teach chirality to our undergrads !
411730
Reposted by @junhyongkim.bsky.social
Heng Li @lh3lh3.bsky.social · 14/01/2026
I am looking for a postdoc to develop high-performance algorithms in computational genomics. Email or DM me if interested. For more information, see hlilab.github.io/vacancies. RTs appreciated!
hlilab.github.io
HLi Lab - Vacancies
Openings
14363
junhyongkim.bsky.social @junhyongkim.bsky.social · 14/01/2026
Since there are so much happenings that I can't breath, ill just have to take it out on this little irritant. What makes a workshop "prestigous" ? Does attending this give trainees prestige? Do organizers give themselves prestige for running it? Can we stop using language like this?
100
junhyongkim.bsky.social @junhyongkim.bsky.social · 08/01/2026
When I first started interviewing, id go to these depts, do the 1 on 1's and then feel bad about the aged scientists who are trying to be relevant. (Ok, not every aged PI, but you know what i am talking about) I promised myself, I am not going to pretend I am relevant. I'll just be outta here.
010
junhyongkim.bsky.social @junhyongkim.bsky.social · 07/01/2026
If y'all would log off, maybe my biosketch will work. 😁
010
Reposted by @junhyongkim.bsky.social
Carl T. Bergstrom @carlbergstrom.com · 30/12/2025
Every week, Nature publishes yet another breathless puff piece about some AI startup, based only unpublished claims from the company and interviewing only those who work there. How can the leading scientific journal publish piece after piece that would make Kevin Roose blush? I think it's that...
21640134
Reposted by @junhyongkim.bsky.social
Russ Fischer @superelastic.bsky.social · 23/12/2025
and the winner of the "2025 best thing on Internet" has just arrived neal.fun/size-of-life/ @carlbergstrom.com
neal.fun
Size of Life
From an amoeba to a blue whale
113656
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
There is a revolution going on with Deep Learning applications in bioscience. However, I do not believe most methods can be evaluated with the standard scientific framework. And, I don't see a solution, just the trendy rush to publish. /end
050
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Models like LLM, also has ambiguous evaluations. But, at least the user has some sense of what is useful or good. The user can clearly see the utility difference of ChatGPT-3 vs 5.x. How would biologists know whether AlphaGenome is practically more useful than, say Enformer? 8/n
110
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Unlike other learning problems, in bio not only do we not have common benchmarks, but the supervised info almost surely has more than 5% error. "Known" information cannot tell us the significance of moderate performance increments. Of course, this is compounded by benchmark selection bias. 7/n
130
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Q3: Is it impactful? If Q1&Q2 is "no or unknown", at least we can examine the results, right? AG claims SOTA performance--as do all DL biology papers. But, how do we know, say, 5% improvement is meaningful? 6/n
120
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
AG has ~0.5B parameters. That's a big space of possible models. Is the AG model novel, not possible otherwise? Could nobody have done something similar? I am sure training/opt was tricky, but how do we reviewers get to evaluate that? This is pretty much true for most bio-DL methods I've seen. 5/n
110
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Q2: Is it a non-trivial solution? Well, some aspects of AlphaGenome (AG) architecture seems both reasonable and expected (coarse-graining, transformer stacks, interaction blocks, etc.). But, if this is a unique and unexpected solution, how would we know? 4/n
110
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Q1: Novel idea? Clearly, a DNA language model is not novel. We can track papers from 90's like doi.org/10.1038/natu... or doi.org/10.1093/nar/... There are even earlier ideas such as this 1984 doi.org/10.1093/nar/... and the classic 1983 "Time warps" book: ISBN 9780201078091. 3/n
doi.org
The language of genes - Nature
Nature - The language of genes
110
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
Note, the following is not a critique of AlphaGenome, per se. It is an impressive feat of engineering. But, I present this example bc I think similar Qs need to be asked for every DL paper in biology: (1) Is the idea/problem novel? (2) Is there a non-trivial solution? (3) Is it impactful? 2/n
110
junhyongkim.bsky.social @junhyongkim.bsky.social · 22/12/2025
I just finished my 28th yr of teaching grad compbio. Following the inevitable trend, a 3rd of the course is now Deep Learning (DL). One activity we did was a deep dive into the AlphaGenome pre-print (www.biorxiv.org/content/10.1...). Question is, how do we evaluate it as a science paper? 1/n
biorxiv.org
AlphaGenome: advancing regulatory variant effect prediction with a unified DNA sequence model
Deep learning models that predict functional genomic measurements from DNA sequence are powerful tools for deciphering the genetic regulatory code. Existing methods trade off between input sequence le...
1204
Reposted by @junhyongkim.bsky.social
Nandita Garud @nanditagarud.bsky.social · 17/12/2025
Grateful to share our paper on gene-specific selective sweeps in human gut microbiomes, now out in Nature! It has been a joy to work with @rwolff.bsky.social, whose insights and hard work made this possible. www.nature.com/articles/s41...
nature.com
Gene-specific selective sweeps are pervasive across human gut microbiomes - Nature
Development and application of the integrated linkage disequilibrium score (iLDS) reveals both selective pressures impacting the human gut microbiome and the mechanisms by which gut bacteria adapt to ...
1014969
junhyongkim.bsky.social @junhyongkim.bsky.social · 13/12/2025
Bless you
010
Reposted by @junhyongkim.bsky.social
Mia Levine @levine-lab.bsky.social · 29/11/2025
Check out this terrific thread by first author, @sungyalin.bsky.social, whose creativity, bravery, exactitude, and perseverance made this project a reality.
072
Reposted by @junhyongkim.bsky.social
Mia Levine @levine-lab.bsky.social · 28/11/2025
How to keep in step when your (protein) partner speeds up… Here we investigated the adaptive remodeling of a protein-protein interaction surface essential for telomere protection. Congrats to whole team! www.science.org/doi/10.1126/...
science.org
Rapid compensatory evolution within a multiprotein complex preserves telomere integrity
Intragenomic conflict with selfish genetic elements spurs adaptive changes in subunits of essential multiprotein complexes. Whether and how these adaptive changes disrupt interactions within such comp...
612064
Reposted by @junhyongkim.bsky.social
Maitreya Dunham @maitreya.bsky.social · 24/11/2025
Open faculty position (Assistant or Associate) in UW Medical Genetics. As you might expect, faculty often end up also interacting with Genome Sciences, so we're hoping for some great prospects! Note the clinical requirements for the position. apply.interfolio.com/176466
apply.interfolio.com
Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio
01930