Arjun Raj
@arjunraj.bsky.social
Just another LLM. Tweets do not necessarily reflect the views of people in my lab or even my own views last week. http://rajlab.seas.upenn.edu https://rajlaboratory.blogspot.com
Reposted by Arjun Raj
The TIG1-high state exists in human lung club and ciliated cells, enriched in IPF patients and tobacco users. Different viruses target different states. Influenza A prefers KRT8-high cells. Arjun Raj @cp-cell.bsky.social
doi.org/10.1016/j.ce... 🧪
3/3
doi.org/10.1016/j.ce... 🧪
3/3
January 20, 2026 at 12:51 PM
The TIG1-high state exists in human lung club and ciliated cells, enriched in IPF patients and tobacco users. Different viruses target different states. Influenza A prefers KRT8-high cells. Arjun Raj @cp-cell.bsky.social
doi.org/10.1016/j.ce... 🧪
3/3
doi.org/10.1016/j.ce... 🧪
3/3
Reposted by Arjun Raj
Reffsin, Miller, Cherry & Raj et al. discover that intrinsic cell states determine which cells become infected with SARS-CoV-2. Using single-cell clone tracing, they identified a TIG1-high state marking highly susceptible cells. What makes these cells vulnerable? 👇🏻
1/3
1/3
January 20, 2026 at 12:50 PM
Reffsin, Miller, Cherry & Raj et al. discover that intrinsic cell states determine which cells become infected with SARS-CoV-2. Using single-cell clone tracing, they identified a TIG1-high state marking highly susceptible cells. What makes these cells vulnerable? 👇🏻
1/3
1/3
Reposted by Arjun Raj
Data and experiments move faster, but theory is what gives discovery direction, via @arjunraj.bsky.social
arjunrajlab.substack.com/p/chance-fav...
arjunrajlab.substack.com/p/chance-fav...
Chance favors the (theoretically) prepared mind
In defense of the role of theory in innovation
arjunrajlab.substack.com
January 19, 2026 at 4:44 PM
Data and experiments move faster, but theory is what gives discovery direction, via @arjunraj.bsky.social
arjunrajlab.substack.com/p/chance-fav...
arjunrajlab.substack.com/p/chance-fav...
Reposted by Arjun Raj
We spend a lot of energy choosing the right research project. But an equally important skill is knowing when to walk away.
Quitting isn’t failure: It’s how you make room for better science, via @arjunraj.bsky.social
Quitting isn’t failure: It’s how you make room for better science, via @arjunraj.bsky.social
January 12, 2026 at 4:44 PM
We spend a lot of energy choosing the right research project. But an equally important skill is knowing when to walk away.
Quitting isn’t failure: It’s how you make room for better science, via @arjunraj.bsky.social
Quitting isn’t failure: It’s how you make room for better science, via @arjunraj.bsky.social
Reposted by Arjun Raj
libertypuzzles.com these are very nice
Liberty Puzzles | Wooden Jigsaw Puzzles
The original wooden jigsaw puzzle maker. Established in 2005, all of our 1000+ puzzles are handmade in Boulder, CO with exquisite quality & craftsmanship. Full of odd-shaped "whimsy" pieces, our woode...
libertypuzzles.com
January 10, 2026 at 8:38 PM
libertypuzzles.com these are very nice
Reposted by Arjun Raj
Anyone have a thought about this weird mol on my hand?
January 10, 2026 at 7:58 PM
Anyone have a thought about this weird mol on my hand?
Reposted by Arjun Raj
I am happy to share that I’ve been awarded tenure here at Columbia University. Alhamdulillah. I thank my wife, parents, mentors, colleagues, friends, and family. I thank my lab members, present and past, for entrusting their careers in my hands and making this a thrilling journey.
July 2, 2025 at 1:43 PM
I am happy to share that I’ve been awarded tenure here at Columbia University. Alhamdulillah. I thank my wife, parents, mentors, colleagues, friends, and family. I thank my lab members, present and past, for entrusting their careers in my hands and making this a thrilling journey.
AI is awesome, and is most definitely a transformative technology.
I’m pretty sure if you say anything positive about AI here you get permanently banned.
January 7, 2026 at 2:29 PM
AI is awesome, and is most definitely a transformative technology.
Reposted by Arjun Raj
To me, the mean loud position here ranges between ignoring the ai and wishing to fully ban it. The latter won't happen (I don't think even regulation is technically feasible), the former is massively unproductive. If scientists refuse to participate in the discussion,it will just be driven by others
January 7, 2026 at 11:18 AM
To me, the mean loud position here ranges between ignoring the ai and wishing to fully ban it. The latter won't happen (I don't think even regulation is technically feasible), the former is massively unproductive. If scientists refuse to participate in the discussion,it will just be driven by others
So… Bluesky for science and X for Claude Code?
January 7, 2026 at 9:58 AM
So… Bluesky for science and X for Claude Code?
Blog post: Chance favors the (theoretically) prepared mind
Data is big, machines are learning, so what good is theory anyway? Isn't most discovery driven by serendipity anyway, with theory mostly a "post-mortem"? I argue that this view discounts the value of theory.
open.substack.com/pub/arjunraj...
Data is big, machines are learning, so what good is theory anyway? Isn't most discovery driven by serendipity anyway, with theory mostly a "post-mortem"? I argue that this view discounts the value of theory.
open.substack.com/pub/arjunraj...
Chance favors the (theoretically) prepared mind
In defense of the role of theory in innovation
open.substack.com
January 4, 2026 at 3:56 PM
Blog post: Chance favors the (theoretically) prepared mind
Data is big, machines are learning, so what good is theory anyway? Isn't most discovery driven by serendipity anyway, with theory mostly a "post-mortem"? I argue that this view discounts the value of theory.
open.substack.com/pub/arjunraj...
Data is big, machines are learning, so what good is theory anyway? Isn't most discovery driven by serendipity anyway, with theory mostly a "post-mortem"? I argue that this view discounts the value of theory.
open.substack.com/pub/arjunraj...
Reposted by Arjun Raj
Beautiful - recommended! Here, @sasolla.bsky.social recaps her decades-long journey from physics to neural networks (working with LeCun & Hopfield) to motor cortex, & and from industry (including Bell Labs) to academia, all driven by curiosity and awe (which flows from her voice). Inspiring!
Episode #36 in #TheoreticalNeurosciencePodcast: On low-dimensional manifolds in motor cortex – with Sara Solla @sasolla.bsky.social
theoreticalneuroscience.no/thn36
Manifold analysis has changed our thinking on how cortex works. One of the pioneers of this modelling approach explains.
theoreticalneuroscience.no/thn36
Manifold analysis has changed our thinking on how cortex works. One of the pioneers of this modelling approach explains.
January 4, 2026 at 11:03 AM
Beautiful - recommended! Here, @sasolla.bsky.social recaps her decades-long journey from physics to neural networks (working with LeCun & Hopfield) to motor cortex, & and from industry (including Bell Labs) to academia, all driven by curiosity and awe (which flows from her voice). Inspiring!
Reposted by Arjun Raj
i've re-implemented PoPE from scratch as part of a model overhaul (w/ Claude Code). i don't have clean ablations yet but it's working. took about two hours to re-write my entire model
i think the PoPE idea is really cool and a great fit for scientific data
i think the PoPE idea is really cool and a great fit for scientific data
They find that RoPE (the positional encoding used in most modern LLMs) has a fundamental flaw. It entangles "what" (content) and "where" (position) information.
They propose PoPE (Polar Coordinate Position Embeddings), which eliminates the what-where.
They propose PoPE (Polar Coordinate Position Embeddings), which eliminates the what-where.
January 2, 2026 at 5:19 AM
i've re-implemented PoPE from scratch as part of a model overhaul (w/ Claude Code). i don't have clean ablations yet but it's working. took about two hours to re-write my entire model
i think the PoPE idea is really cool and a great fit for scientific data
i think the PoPE idea is really cool and a great fit for scientific data
Reposted by Arjun Raj
2025 was the year we stopped debating whether paying peer reviewers might work and started showing that it does.
December 31, 2025 at 10:13 PM
2025 was the year we stopped debating whether paying peer reviewers might work and started showing that it does.
Reposted by Arjun Raj
Agreed. The sunken cost fallacy is real. What I tend to do is to put projects on the shelf and once in a while look at them again wondering whether new technology or data would allow us to make progress. Emotionally it feels less like quitting and more as postponement
December 30, 2025 at 8:07 PM
Agreed. The sunken cost fallacy is real. What I tend to do is to put projects on the shelf and once in a while look at them again wondering whether new technology or data would allow us to make progress. Emotionally it feels less like quitting and more as postponement
Reposted by Arjun Raj
Great advice received as a postdoc: Do the easiest thing.
I was 3 years into my postdoc and juggling projects w/nothing concrete = career crisis
I went to my advisor Paul Sternberg for help. He asked what was going on & what mutant had the strongest phenotype.
"Clone the easiest one."
It worked.
I was 3 years into my postdoc and juggling projects w/nothing concrete = career crisis
I went to my advisor Paul Sternberg for help. He asked what was going on & what mutant had the strongest phenotype.
"Clone the easiest one."
It worked.
Blog post: Just quit
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Just quit
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com
December 30, 2025 at 5:06 PM
Great advice received as a postdoc: Do the easiest thing.
I was 3 years into my postdoc and juggling projects w/nothing concrete = career crisis
I went to my advisor Paul Sternberg for help. He asked what was going on & what mutant had the strongest phenotype.
"Clone the easiest one."
It worked.
I was 3 years into my postdoc and juggling projects w/nothing concrete = career crisis
I went to my advisor Paul Sternberg for help. He asked what was going on & what mutant had the strongest phenotype.
"Clone the easiest one."
It worked.
Bad antibodies are such a huge problem. It’s not just the money wasted on buying them. It’s the amount of time spent trying to validate them.
Many articles reported using an antibody ... that didn’t even bind the key protein in Laflamme’s testing. Those articles had been cited over 3,000 times.
A key question is:
◼️ 𝐂𝐚𝐧 𝐩𝐫𝐨𝐭𝐞𝐢𝐧 𝐛𝐢𝐧𝐝𝐞𝐫𝐬 𝐚𝐜𝐡𝐢𝐞𝐯𝐞 𝐡𝐢𝐠𝐡 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜𝐢𝐭𝐲 𝐟𝐨𝐫 𝐞𝐯𝐞𝐫𝐲 𝐞𝐩𝐢𝐭𝐨𝐩𝐞 & 𝐩𝐫𝐨𝐭𝐞𝐨𝐟𝐨𝐫𝐦?
A key question is:
◼️ 𝐂𝐚𝐧 𝐩𝐫𝐨𝐭𝐞𝐢𝐧 𝐛𝐢𝐧𝐝𝐞𝐫𝐬 𝐚𝐜𝐡𝐢𝐞𝐯𝐞 𝐡𝐢𝐠𝐡 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜𝐢𝐭𝐲 𝐟𝐨𝐫 𝐞𝐯𝐞𝐫𝐲 𝐞𝐩𝐢𝐭𝐨𝐩𝐞 & 𝐩𝐫𝐨𝐭𝐞𝐨𝐟𝐨𝐫𝐦?
December 30, 2025 at 3:04 PM
Bad antibodies are such a huge problem. It’s not just the money wasted on buying them. It’s the amount of time spent trying to validate them.
Blog post: Just quit
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Just quit
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com
December 30, 2025 at 3:02 PM
Blog post: Just quit
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Quitting projects in science is hard, but we should be doing a lot more of it.
open.substack.com/pub/arjunraj...
Reposted by Arjun Raj
Our preprint "Predictive design of tissue-specific mammalian enhancers that function in vivo in the mouse embryo" is on bioRxiv: www.biorxiv.org/content/10.6... . Amazing collaboration by @shenzhichen1999.bsky.social, Vincent Loubiere (@impvienna.bsky.social,@viennabiocenter.bsky.social),... (1/2)
Predictive design of tissue-specific mammalian enhancers that function in vivo in the mouse embryo
Enhancers control tissue-specific gene expression across metazoans. Although deep learning has enabled enhancer prediction and design in mammalian cell lines and invertebrate systems, it remains uncle...
www.biorxiv.org
December 24, 2025 at 3:06 PM
Our preprint "Predictive design of tissue-specific mammalian enhancers that function in vivo in the mouse embryo" is on bioRxiv: www.biorxiv.org/content/10.6... . Amazing collaboration by @shenzhichen1999.bsky.social, Vincent Loubiere (@impvienna.bsky.social,@viennabiocenter.bsky.social),... (1/2)
Reposted by Arjun Raj
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
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
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...
www.biorxiv.org
December 22, 2025 at 7:10 PM
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
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
Reposted by Arjun Raj
A new paper out in @natmethods.nature.com introduces NimbusImage, a cloud-computing platform for image analysis ft. William Niu (SAS), Jingxin Li (@cambupenn.bsky.social), @aofarrell.bsky.social (Bioengineering) & @arjunraj.bsky.social (@penngenetics.bsky.social) www.nature.com/articles/s41...
December 19, 2025 at 7:26 PM
A new paper out in @natmethods.nature.com introduces NimbusImage, a cloud-computing platform for image analysis ft. William Niu (SAS), Jingxin Li (@cambupenn.bsky.social), @aofarrell.bsky.social (Bioengineering) & @arjunraj.bsky.social (@penngenetics.bsky.social) www.nature.com/articles/s41...
Reposted by Arjun Raj
We designed SpaceBar to be as easy to use as possible:
Cells are labeled via lentivirus transduction, and barcode detection works with any standard imaging-based spatial transcriptomics workflow.
w/ Yael Heyman and @arjunraj.bsky.social
Check out the thread from the preprint for more info!
Cells are labeled via lentivirus transduction, and barcode detection works with any standard imaging-based spatial transcriptomics workflow.
w/ Yael Heyman and @arjunraj.bsky.social
Check out the thread from the preprint for more info!
Excited to share SpaceBar - our new method for labeling and detecting clones with imaging-based spatial transcriptomics platforms! w/ Yael Heyman and @arjunraj.bsky.social www.biorxiv.org/content/10.1... 🧵
December 18, 2025 at 7:55 PM
We designed SpaceBar to be as easy to use as possible:
Cells are labeled via lentivirus transduction, and barcode detection works with any standard imaging-based spatial transcriptomics workflow.
w/ Yael Heyman and @arjunraj.bsky.social
Check out the thread from the preprint for more info!
Cells are labeled via lentivirus transduction, and barcode detection works with any standard imaging-based spatial transcriptomics workflow.
w/ Yael Heyman and @arjunraj.bsky.social
Check out the thread from the preprint for more info!
Reposted by Arjun Raj
Stoked that this paper, summarizing a vast series of experiments from past and present members of the team is out. A great way to end this year.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Whole body regeneration deploys a rewired embryonic gene regulatory network logic - Nature Communications
To what extent regeneration recapitulates embryonic development is a longstanding question. Here, they show that embryonic gene modules are re-used, rewired, and interconnected to specific injury-indu...
www.nature.com
December 17, 2025 at 3:57 PM
Stoked that this paper, summarizing a vast series of experiments from past and present members of the team is out. A great way to end this year.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
I just pinned a photo of a post-it to one of our lab Slack channels. Is this what they mean by "tech stack"?
December 17, 2025 at 5:57 PM
I just pinned a photo of a post-it to one of our lab Slack channels. Is this what they mean by "tech stack"?
Reposted by Arjun Raj
Absolutely thrilled to share the latest work from my lab focused on the variation and evolution of human centromeres among global populations! We assembled 2,110 human centromeres, identifying 226 new major haplotypes and 1,870 α-satellite HOR variants. www.biorxiv.org/content/10.6...
December 16, 2025 at 4:06 PM
Absolutely thrilled to share the latest work from my lab focused on the variation and evolution of human centromeres among global populations! We assembled 2,110 human centromeres, identifying 226 new major haplotypes and 1,870 α-satellite HOR variants. www.biorxiv.org/content/10.6...