Steven ten Holder

1.4K posts

Steven ten Holder banner
Steven ten Holder

Steven ten Holder

@steventen

building zeroshot bio // prev. co-founded @acornbiolabs, YC 2016

Vancouver, British Columbia 가입일 Eylül 2012
3.1K 팔로잉1.8K 팔로워
evo-devo
evo-devo@Xiaojie_Qiu·
Today in Nature Methods, we propose a bold new vision for biology: Virtual Embryos 🧬🤖 By integrating single-cell and spatial genomics with AI, we can build digital twins of embryogenesis—moving beyond virtual cells to predict cell growth, division, migration, state transitions, and morphogenesis, from genes → cells → organs -> organ systems and whole embryo, in a fully 4D spatiotemporal framework. Congenital defects affect ~1 in 33 births in the U.S., costing ~$23B annually. This approach could enable truly predictive biology and in silico experimentation to diagnose, prevent, and treat developmental diseases—transforming medicine and improving outcomes for future generations. If this excites you, come build with us—together with Emily Fox, James Zou (@james_y_zou), and Marinka Zitnik (@marinkazitnik). We also welcome industrial and philanthropic support for this ambitious vision.
evo-devo tweet mediaevo-devo tweet media
English
4
61
274
20.7K
Perturb.ai
Perturb.ai@perturbai_tx·
Our platform, the engine behind the world’s largest in vivo CRISPR atlas, was featured in a keynote at @NVIDIAGTC this week. Collaborating with @NVIDIAHealth and others enabled us to analyze a dataset of 8 million brain-wide cells with CRISPR edits, establishing a new category of biological data: organism-level, circuit-resolved, causal genomics. We are utilizing our in vivo CRISPR platform and causal AI models to develop best-in-class therapeutics for complex metabolic and chronic diseases. Read more about our collaboration here: #bionemo" target="_blank" rel="nofollow noopener">blogs.nvidia.com/blog/gtc-2026-… #AI #CRISPR #DrugDevelopment #Innovation #Genomics
Perturb.ai tweet media
English
1
2
9
586
Sauers (in Berkeley / SF)
Woah--Claude Opus 4.6 in Claude chat is much better than GPT-5.4 thinking for scientific figures. This is new since the last time I tried
English
3
1
71
3.3K
Lisa Tanh
Lisa Tanh@LisaLi_T·
If you're in Vancouver and in the healthcare / health tech space, my friends are hosting a meetup for builders, investors, and clinicians! luma.com/ya2d5hkd?tk=Ag…
English
1
1
6
177
Sauers (in Berkeley / SF)
My project to build disease prediction models, formalize (and prove new) genomic theory in Lean 4, discover powerful new methods for computational biology, and create the best generalized additive model engine, has been accepted into the @AnthropicAI Open Source Program!
Sauers (in Berkeley / SF) tweet media
English
27
12
348
20.6K
Seth Bannon
Seth Bannon@sethbannon·
A shocking number of synthetic biologists got into it because of jurassic park.
English
19
9
153
159.8K
Ron Alfa
Ron Alfa@Ronalfa·
Imagine actually claiming you are the openAI of bio at seed stage.
English
9
0
56
10.2K
Steven ten Holder
Steven ten Holder@steventen·
@DylanoA4 I wonder whether working through the puzzles by simulating them with claude as you go would enrich the reading experience..
English
0
0
2
482
Ron Alfa
Ron Alfa@Ronalfa·
@steventen I think we can approach this but I wouldn’t do scRNA seq.
English
1
0
1
54
Ron Alfa
Ron Alfa@Ronalfa·
Our oddly contrarian thesis: human bio foundation models need to be trained on actual human data.
English
6
0
29
2.5K
Steven ten Holder
Steven ten Holder@steventen·
@Ronalfa honestly way better, but the ideal is unattainable -- full human-body scRNA seq across time points.
English
1
0
1
52
Ron Alfa
Ron Alfa@Ronalfa·
@steventen Tissues from patients. Cell culture is not human.
English
1
0
0
53
Saganism
Saganism@Saganismm·
"We are like butterflies who flutter for a day and think it is forever." — Carl Sagan
Saganism tweet media
English
28
207
923
22.6K
Arc Institute
Arc Institute@arcinstitute·
It's an exciting time to be at Arc. We recently welcomed a new CSO, announced a major partnership with Tahoe Therapeutics and Biohub, and are continuing to open source models and techniques to push forward progress on virtual cells.
English
3
4
53
6.2K
Steven ten Holder
Steven ten Holder@steventen·
@wildtypehuman It does feel like we're just a few orders of magnitude away from comp bio crossing into a new paradigm of runaway capability. Currently, brilliant people working with insufficient data spinning their wheels is what makes @Andrei_Tarkhov feel the way he does.
English
0
0
1
221
Jake P. Taylor-King
Jake P. Taylor-King@wildtypehuman·
How to show you don’t know what comp bio is. In fact, comp bio is probably the *most* employable skillset at the moment, especially at the research end. Why? Well new ‘omic technologies are coming out constantly and we are still learning how to use them. Some bioinformatics jobs may go due to pipelines becoming routine (eg WGS etc), but comp bio writ large? No way.
Andrei Tarkhov, PhD@Andrei_Tarkhov

compbio is about to die. A popular general-purpose AI assistant is already able to write a whole analysis pipeline starting from a simple prompt and a link to the data. Surely, I can do it 10 times faster and using 10 times fewer lines of code, but still “if it works, it works”

English
7
6
71
9.7K
Steven ten Holder
Steven ten Holder@steventen·
@FyruzOne @Andrei_Tarkhov I guess I’m implying scale and turnaround time is itself a threshold we cross into new discovery territory. I see your point tho — it’s not directly new categories of capability.. yet
English
0
0
1
25
fyruz
fyruz@FyruzOne·
@steventen @Andrei_Tarkhov more capability for biocomp would be if at the same time we develop more new stuff because of this efficiency but this isn't necessarily the case. We might just be able to do complete NGS pipelines in 1 day instead of 7 without doing anything new and meaningful on the remaining 6
English
1
0
1
31
Andrei Tarkhov, PhD
Andrei Tarkhov, PhD@Andrei_Tarkhov·
compbio is about to die. A popular general-purpose AI assistant is already able to write a whole analysis pipeline starting from a simple prompt and a link to the data. Surely, I can do it 10 times faster and using 10 times fewer lines of code, but still “if it works, it works”
English
18
11
158
43.4K