@lucas_nivon profile picture
Lucas Nivon ๐Ÿ‡บ๐Ÿ‡ธ @lucas_nivon

CEO and Co-Founder at Cyrus Bio. Openfold AI consortium co founder. Computational biologic drug discovery. Immunogenicity, MHCII, Auto-immune. New Yorker

Joined March 2013
  • Posts 975
  • Following 842
  • Followers 739
32 Photos and videos
Very beautiful benchmarking of free energy โ€” public datasets and the. Private data from 15 (!) pharmas. More error in the presumably more difficult private molecules. Exemplary methodology for real world comparison
JCIM & JCTC Journals @JCIM_JCTC
Jul 5
Large-Scale Collaborative Assessment of Binding Free Energy Calculations for Drug Discovery Using OpenFE pubs.acs.org/doi/10.1021/acsโ€ฆ #JCIM Vol66 Issue11 #compchem
2
9
2,118
Very cool Iโ€™ve wondered about mole rats lifespan. Cue for a tech bro newco selling you 28s rRNA injections as a subscription service with concierge rRNA monitoring for $100k/mth
Misha Kudryashev ๐Ÿ”ฌ โ„๐Ÿ•Š๏ธ @MishaKudryashev
21h
New paper from the group! We were curious if the longevity of naked mole-rats has anything to do with their protein synthesis machinery. It has been shown that the ribosomal 28S is split into 3 subunits Our cryo-EM study of the NMR ribosomes, take a look: nature.com/articles/s41467-0โ€ฆ
1
7
2,676
Great post on difficult business path for #ai antibodies, mirror image of my piece on AI designed non antibody biologics linkedin.com/pulse/non-antibโ€ฆ

Non-antibody biologics are the low-hanging fruit of AI-driven biologics discovery

Cyrus focuses on developing new non-antibody biologics with improved PK/PD, immunogenicity, and specificity over natural proteins. We are often asked why we do not focus on antibodies, since most...

linkedin.com
Carlos Outeiral
 @c_outeiral
Jul 2
Everyone's talking about AI-designed antibodies. Having sat through several such "revolutions," I was wondering, is this the real one, or just another flop? So I wrote a long-form essay on whether it holds up. Spoiler: the science checks out, not so sure about the business.
1
3
20
3,464
Congrats to @AnthropicAI on Claude science โ€” @open_fold is available immediately via the @nvidia bionemoskills integration! blogs.nvidia.com/blog/claudeโ€ฆ

NVIDIA BioNeMo Agent Toolkit Brings Accelerated AI to Life Sciences Researchers in Claude Science

Anthropicโ€™s Claude Science, an AI workbench for science research, lets scientists use agents to build workflows harnessing NVIDIA BioNeMo models, libraries and accelerated workflows.

blogs.nvidia.com
2
14
690
And if you have your own data you can fine tune openfold3 yourself and reach to to @open_fold for membership and professional help with every step of the way! This is #sovereign #ai that no other company controls!
1
57
Or you can host on your hardware or private cloud eg with tools like @fastmcp . Unlike proprietary models like Boltz or Chai you can run your own version of the latest models without putting your valuable data on someone elseโ€™s server
41
Lucas Nivon ๐Ÿ‡บ๐Ÿ‡ธ reposted
Science in Seattle
 @science_seattle
Jun 25
Nate Corley and Dr. Frank DiMaio at the @UWproteindesign have partnered with NVIDIA to explore custom computing architectures built specifically for biological design challenges. See how this could unlock new innovations in medicine and biotech: bit.ly/4uXwAEh
4
14
657
Such tragic deaths caused by USAID cuts, highlighted in detail by Dr. Gawande
Atul Gawande
 @Atul_Gawande
Jun 25
Independent analyses estimate that your actions to dismantle USAID and drastically reduce lifesaving foreign aid have already killed 700,000 people. Here are some of those people. ๐Ÿงตx.com/elonmusk/status/206944โ€ฆ
3
188
Very exciting to see a new funding concept in infectious disease, corporate purchase commitments used to fund difficult science!
Charlie Petty
 @incredutility
Jun 24
I joined @Stripe in January to build a new philanthropic initiative with @nanransohoff. Today weโ€™re announcing Intercept, a $500m innovation fund focused on radically reducing the burden of respiratory viruses. Iโ€™ve spent the past decade investing in vaccines and drugs for infectious disease, and this is the most exciting thing Iโ€™ve ever been a part of. Thrilled to be working with @Stripe, @AnthropicAI, @TheFluLab, @FoundationOAI, @coeff_giving and individuals from Jane Street. We've raised the capital and now itโ€™s time to get to work.
1
6
740
My daughterโ€™s #urbanism solution: adults disguised as kids start a chain of lemonade stands that also sell coffee. On corners all around the burbs. Legal mixed use everywhere!
136
Can we have an authentic discussion about Seattle new growth? It goes back to overall lack of mega cap privates in tech and biotech. Not taxes (see SF BOS). Not crime or disorder ( SF). Big capital? Ambition? In biotech, senior talent?
1
190
Bio AI Nobel prize leaderboard: Openfold 0.5 Nobels, deepmind/Isomorphic 0.25, Chai 0, Boltz 0, all other at zero. Join Openfold, the sustainable future of protein AI!
John Jumper
 @JohnJumperSci
Jun 19
A bit of news: After nearly 9 years, I have decided to leave Google DeepMind and join Anthropic (after taking some time to recharge). I am incredibly grateful for my time at GDM. @demishassabis took a real chance letting me lead the AlphaFold team just six months after finishing my PhD, and the entire GDM team taught me so much about how to do great science. GDM is a special place, and Iโ€™ll still be excited to hear about what amazing things they discover next.
1
7
687
What was his last Seattle angel investment? I see zero in pitchbook. Seattle needs to believe in itself. We need Seattle capital to invest locally, I have amazing local angels but institutional capital is nyc bos etc
GeekWire
 @geekwire
Jun 18
Loss of another Seattle-area billionaire? Valveโ€™s Gabe Newell is reported buyer of Florida estate geekwire.com/2026/loss-of-anโ€ฆ
1
282
One of the key results of OpenBind was that OpenFold and Protenix are ahead of Alphafold3 on this truly blind test set for protein/ligand. Far ahead of Chai and Boltz2. On top of that Openfold is open source open weight. Protenix trained on a later cutoff. openbind.uk/news/blog-openbiโ€ฆ
3
21
148
9,513
The huge surprise was AF3, maybe over tuned for runs n poses? Iโ€™ve never seen an independent bench where Chai or Boltz beat AF3 or OF3p2. I donโ€™t understand the enthusiasm for those models other than that they were earlier than OF3
2
383
Ah realized here no Chai, but internal tests not close for protein ligand. Protein protein looks better for Chai. Maybe openfold needs to release more hype videos?
1
1
499
I hope we see some clear analysis from physicians about this, but this wrote up asks the right questions
Hank Green
 @hankgreen
Jun 18
x.com/i/article/206746144069โ€ฆ
1
225
With OF3p2 and other OF models your scientists can fine tune, run inference directly, add new attention heads (we do that @cyrusbiotech ) and continue innovating. All of that is the essence of new science. Not possible with proprietary models like IsoDDE, Boltz, Chai, etc.
1
1
330
When we started @open_fold we considered various models and realized that only a true 501c3 funded by a huge number of Pharmas, like Linux for code, could maintain leading performance and open source availability for the long haul. Openfold will continue to release for everyone!
1
2
235
I am glad that companies like Boltz are continuing to innovate in protein AI, but this release shows why @open_fold and open source open weights bio #ai is the future of biotech. Eventually VC funded AI will get paywalled and limit innovation. Join @open_fold today!
Gabriele Corso @GabriCorso
Jun 16
Big news from Boltz - our biggest update yet! ๐Ÿš€ Today weโ€™re releasing two new state-of-the-art models for protein and small molecule design with extensive wet lab validation and a new API to run all of our models on scalable GPUs wherever you (or your agents) work! ๐Ÿ”ฅ
0:11
1
1
36
2,670
Load more