AI x drug discovery. Currently working on something new. Previously @Dyno_Tx @Princeton @JohnsHopkins
New York, NY
Joined March 2012
- Tweets118
- Following1.5K
- Followers359
- Likes1.3K
Farhan Damani retweeted
We’ve now licensed our second AI-designed capsid, this time targeting muscle (after CNS). The only benchmark that counts is whether it works in the body. Doing it again, in a different organ, shows the power of our platform. Huge congrats to the team and excited to see this help patients.
We’re proud to announce that @AstellasUS has exercised its $15M option to license a Dyno AI-designed adeno-associated virus (AAV) capsid for skeletal muscle-targeted gene delivery, building on our 2021 collaboration agreement.
This marks our second capsid license to a pharmaceutical partner, and it makes Dyno the first company to license AI-designed AAV capsids for both central nervous system (CNS) and muscle gene therapies.
This milestone reflects years of platform development, validating our conviction that AI can solve the gene delivery challenge for therapeutic developers and ultimately create many good treatment options for patients.
Learn more: businesswire.com/news/home/2…
Farhan Damani retweeted
Hub-and-Spoke Biotechs: Nimbus Therapeutics
I wrote about how Nimbus's fundamental business model innovation enabled them to make blockbuster drug deals—without selling their entire company.
Here's how they built an "evergreen" R&D engine.
centuryofbio.com/p/nimbus
Farhan Damani retweeted
On behalf of @Dyno_Tx's ML team, we are so excited for the work we are going to do with @NVIDIAAI. Special thanks to @ReneeYao1 , @anthonycosta (among many others), and of course the visionary Jensen Huang who sees the size of the opportunity at the intersection of #AI and biology. 🚀
We're excited to announce our collaboration with @nvidia. Working with the world’s leading computing company and building within NVIDIA's BioNeMo will accelerate Dyno's AI-powered sequence design and bring us all closer to helping patients with transformative new medicines.
Farhan Damani retweeted
David Brookes* took the main stage @SynBioBeta and gave a brief overview of our collective work @Dyno_Tx today, showing (1) How we can venture deep into sequence space with our AI models, away from the evolutionary trajectory, finding protein sequences that are helpful to humans
David Brooke of Dyno Therapeutics (@Dyno_Tx) reveals how we don’t need to rely on evolution to help sequence diversity. They are developing their own ML technology to move beyond the natural sequences.
#synbiobeta2024 #synbiobeta #synbio #syntheticbiology
Tomorrow I will be giving a talk at #MLCB2023 focused on our recent work at @Dyno_Tx studying distribution shift in ML-guided sequence design. The talk will be live-streamed on YouTube. Our team will also be presenting two posters -- see below for details!
Dyno is heading to #MLCB2023 this Thursday + Friday to share our latest advancements in Machine Learning methods for AAV design and gene therapy! On Nov 30th from 2:35-2:55 PST we’re giving a talk, then at 4:50-6:30 PST we’ll be presenting two posters. Be sure to catch us there!
Farhan Damani retweeted
Amazing presentation by @hayleystrasburg from @lab_schaefer at @gliameeting euroglia meeting in Berlin. Could not be more proud.
Farhan Damani retweeted
Today at #ASGCT23, we shared data on a breakthrough CNS-targeted AAV capsid that dramatically improves pan-brain neuronal transduction & liver detargeting in NHPs at a low IV dose. Learn more at dynotx.com/bcap1/
Farhan Damani retweeted
Interested in learning more about Dyno’s AAV capsids with >100X improved transduction to retina and CNS? Schedule a 1:1 w/ us at BIO 2023 through the One-on-One Partnering portal - see you next week in Boston! #BIO2023
Farhan Damani retweeted
1/ [Tweetorial 🧵] Excited to share some theoretical advances from our team @Dyno_Tx led by David H Brookes, building on his graduate work on exploiting the sparsity of fitness functions in the Walsh-Hadamard/Fourier space for learning and design. arxiv.org/abs/2305.03136
Excited to be at NeurIPS this year along with many colleagues from @Dyno_Tx! Looking forward to chatting with people about ML for sequence design, model-based optimization, and offline RL.
At @Dyno_Tx we are working on building the future of gene therapy with ML guided design. Come work with us! We have lots of open positions in ML and engineering.
Farhan Damani retweeted
Dyno announces our partnership with Astellas, valued at up to $1.6 billion, for developing next-gen AAV gene therapy vectors for skeletal and cardiac muscle diseases. bit.ly/3xHDP7j
Farhan Damani retweeted
We (@Dyno_Tx) just published post 1/3 in our new introduction to AAV blog post series. This post covers AAV biology with the goal of setting the stage for post 2 (on AAV engineering) and post 3 (on Dyno's workflow).
dynotx.com/introduction-to-a…
Farhan Damani retweeted
Can we safely deliver better gene therapies for more diseases? Founder/CEO @ekelsic joins me to discuss how @Dyno_Tx is using AI to reshape the gene therapy landscape 🧬
w/ @vintweeta @vijaypande @pmarca @julesyoo @a16z
Mon May 10 5PM PDT @joinclubhouse. joinclubhouse.com/event/xjlo…
Farhan Damani retweeted
@Dyno_Tx 's $100M Series A led by @a16z, welcome to our team of #AAViators @JorgeCondeBio + @cascapital , @GVteam, @obviousvc & @Lux_Capital. We will use these funds to solve in vivo gene delivery towards helping more patients benefit from #genetherapy!
Dyno Therapeutics Closes $100 Million Series A Financing Led by Andreessen Horowitz to Accelerate AI‑powered Gene Therapy Platform.
dynotx.com/dyno-closes-100-m…
I joined Dyno because of their deep-seated belief in ML-guided design as a powerful platform to help solve the delivery problem in gene therapy. A year later and we're making rapid progress. Exciting times ahead!
Dyno Therapeutics Closes $100 Million Series A Financing Led by Andreessen Horowitz to Accelerate AI‑powered Gene Therapy Platform.
dynotx.com/dyno-closes-100-m…
Farhan Damani retweeted
1/As a side-kick to our bigger supervised modeling project that recently came out, we also tried using the VAE model for proteins I had worked on in 2017 to design 3750 sequences. Here are the results: biorxiv.org/content/10.1101/…
