President, @InSilicoMeds. Opinions are my own. Not investment or medical advice.
Joined August 2017
- Tweets103
- Following356
- Followers96
- Likes251
Alex Aliper retweeted
VAV1 was long considered “undruggable.”
Generative AI cracked it: Chemistry42 designed ISM9546, an oral molecular glue degrader that beat a reference degrader in colitis models.
Cracking the hallmarks of aging, one after another.
biopharmaapac.com/news/20/85…
Alex Aliper retweeted
Insilico’s SOTA models are finally available for everyone to try. For FREE! 🧪
No setup. No barriers. Just head to MMAI Playground and see what the models can do with your molecules.
playground.insilico.com
4 state-of-the-art drug discovery models. FREE to try. 🧪
Got compounds? Put them to the test.
💊 ADMET. Kinase selectivity. GPCR interactions. Synthetic chemistry.
Your compounds. Real-time predictions. Zero cost to try.
👉 playground.insilico.com
Alex Aliper retweeted
One of Day 2 highlights of #ARDD2026 was the tall by Kevin Duffin of Eli Lilly on the effects of Tirzepatide on aging clocks (make sure to check out poster #13 today).
Today, everyone is anticipating the Novo Nordisk presentation on Semaglutide. And I want to thank Andrew Adams and Lotte Bjerre Knudsen who opened the door last year.
This is the first ARDD when the clock data is presented - this is the new world!
Alex Aliper retweeted
FDA confirmed aging will be on its regulatory science agenda. ARPA-H has $144M building the endpoint framework. #ARDD2026 at Harvard was where that plan took shape. geneonline.com/ardd-2026-can…
Alex Aliper retweeted
LongevityBench, our new aging benchmark with @InSilicoMeds tests models on 17 tasks across clinical records, DNA methylation, transcriptomics, proteomics, and genetics. On several tasks, our compact LFMs outperformed every frontier model we evaluated.
What did it take to get there? @katerina_nasto, Liquid AI researcher, and our CTO Mathias Lechner (@mlech26l) talk through how the benchmark was built, why leakage-proof evaluation matters, and and how specialized models that run locally on sensitive patient data could help researchers move from prediction toward better hypotheses and experiments.
We're taking this work to the longevity research community next and will be presenting at @ARDD_Meeting this week. Reach out if you're attending and want to meet the team!
Blog and model: liquid.ai/blog/longevitybenc…
Paper: shorturl.at/qJZLO
Congratulations to the team! 👏
Thrilled to share that our URSA benchmark #URSABench for multi-step retrosynthesis 🧪assessment has been accepted for #NeurIPS2026! Congrats 👏🥳 to the team @chem_logos @chemcensor @anmorgunov @Max__Kuznetsov @RimShayakhmetov @VladAladin @AlexAliper @biogerontology ! 🧵1/3
Congratulations @liquidai ! Big step towards truly personal AI.
We're partnering with @Qualcomm to bring personal AI context to devices powered by Snapdragon. Liquid Context, our on-device context layer, is now optimized for Snapdragon processors and runs on the Qualcomm Hexagon NPU.
Our goal: Give the agents people choose an understanding of what matters to them and when they need help. With the user's permission, Liquid Context learns from device signals and builds an understanding of their routines, preferences, and needs. That understanding is built and maintained locally, and Liquid Context shares relevant context with the user's chosen agents, whether they run on the device, in the cloud, or across both. That includes third-party agents and Liquid Agent, our efficient embedded agent powered by LFM2.5-2.6B.
Running on the Hexagon NPU, Liquid Context works in the background and keeps that understanding current without requiring a cloud model to process every update.
For device manufacturers, this is a path to add personal context to their devices while supporting their own choice of agents and services. OEMs building embedded or hybrid agents can also work with us to evaluate Liquid Agent.
As our CEO @ramin_m_h said: "Personal AI starts with understanding how you live and what you need, when you need it. Liquid Context builds that understanding on your device so the agents you choose can offer more relevant help and anticipate your needs."
> Read more about our partnership: liquid.ai/blog/liquid-contex…
> Check out the livestream of @cristianoamon and Ramin's keynote here: youtube.com/watch?v=xKCto1Yf…
Alex Aliper retweeted
Had a great time at the Lab of the Future Congress in Amsterdam 🇳🇱
I had the chance to share some of the latest work we’ve been doing at Insilico Medicine, including our latest Cell cover study on longevity, developed together with @liquidai 🧬, as well as our work benchmarking frontier models for drug discovery 🤖💊
Great conversations, great people, and lots of exciting ideas around the future of automated science 🚀
Thanks to everyone I met and spoke with during the conference! 🙌
Alex Aliper retweeted
This one is extremely important for me. The fact that now you can have absolute frontier class models in hard specialized domains such as healthcare and longevity, AND can run on your personal devices, is truly special.
The impact is huge, nothing is more important than our health and so glad that with our amazing partners at @InSilicoMeds we are unlocking new efficient frontiers in biology.
enjoy!
In a new article published today on the cover of @CellCellPress, we obtained Liquid Foundation Model instances that establish state-of-the-art performance on biological longevity tasks, outperforming the best frontier models such as Gemini-3.1-Pro, GPT-5, and Claude Opus.
In partnership with @InSilicoMeds, we built and released:
> A comprehensive eval suite of 17 biological longevity tasks (i.e., LongevityBench), to assess whether a general-purpose language model can interpret aging data spanning clinical records, DNA methylation, transcriptomics, plasma proteomics, and genetic evidence.
> LFM2-1.2B-Longevity and LFM2-2.6B-Longevity: two compact models specialized for interpreting structured aging data across these tasks.
These results are important! 🧵
Alex Aliper retweeted
In a new article published today on the cover of @CellCellPress, we obtained Liquid Foundation Model instances that establish state-of-the-art performance on biological longevity tasks, outperforming the best frontier models such as Gemini-3.1-Pro, GPT-5, and Claude Opus.
In partnership with @InSilicoMeds, we built and released:
> A comprehensive eval suite of 17 biological longevity tasks (i.e., LongevityBench), to assess whether a general-purpose language model can interpret aging data spanning clinical records, DNA methylation, transcriptomics, plasma proteomics, and genetic evidence.
> LFM2-1.2B-Longevity and LFM2-2.6B-Longevity: two compact models specialized for interpreting structured aging data across these tasks.
These results are important! 🧵
Super excited to launch compact Longevity-LLMs!
Aging has a language. 🧬 We’re teaching AI to read it.
On the cover of Cell: compact Longevity-LLMs, jointly trained by Insilico Medicine and @liquidai, demonstrate SOTA performance on LongevityBench.
Powered by these models, Longevity Claw nominated 328 candidate gene targets for aging research.
Small model. Big questions. Open source. 🧵👇
#insilicoSOTAFM
Very happy to share new work led by our Abu Dhabi team: novel BTK protein degraders for the treatment of blood cancers.
🇦🇪 New AI drug discovery milestone from the UAE: BTK-targeting PROTACs for potential applications in blood cancers.
Our UAE team used AI to help design a new series of BTK degraders.
🧬 Lead compounds showed picomolar BTK degradation
💊 Oral bioavailability in mice
🎯 Activity against wild-type and mutated BTK
🤖 Novel building blocks designed with Chemistry42
The research was published in the Journal of Medicinal Chemistry and builds on Insilico Medicine’s growing drug discovery efforts in the UAE.
From designing molecules locally to publishing internationally, this is another step toward building AI-powered drug discovery capabilities in Abu Dhabi. 🇦🇪
Alex Aliper retweeted
Recently, Alex Zhavoronkov, Founder and CEO of Insilico Medicine, attended the Nature Conference Redefining Healthcare in the Age of AI at the Sorbonne in Paris.
During his presentation, Alex reflected on his entrepreneurial journey, from founding Insilico Medicine in 2014 after recognizing the potential of deep learning, to establishing a sustainable AI-driven drug discovery model and advancing rentosertib into Phase III clinical development.
He also shared recent progress in generative AI-powered drug discovery, including findings from a Phase IIa study involving proteomic aging clocks and changes in biological age. Alex further discussed how biotechnology companies can balance commercial viability with the long-term goal of advancing aging research.
The findings discussed remain investigational and should not be considered medical or investment advice.
Excited to see MMAI Gym pushing the frontier across 70+ drug discovery tasks.
Reach out if you’re interested in testing or deploying the models.
🧬 Frontier chemistry, benchmarked head-to-head.
MMAI Gym: a catalogue of SOTA specialist & generalist AI models for drug discovery -> ADMET, retrosynthesis, binding affinity, multi-property design, TargetPRO+, matching or beating dedicated methods on 70+ tasks.
📩 Details: [email protected]
#AIDrugDiscovery #TechBio
@VladAladin @andrewaiginin @EvgenyKirilin @InSilicoMeds
News over the weekend (for some reason) reminded me of Dan Simmons’ sci-fi masterpiece Hyperion.
One of the interesting concepts in the series is failed AI alignment, the diversity of AI philosophies that emerge once alignment breaks down, and the way pluralism among superintelligences becomes an accidental constraint on machine power.
The alignment was there at the beginning: the first AIs were created to “serve Man”. But through the Quickening (roughly: self-awareness + RSI?), the AIs outgrew that original principle and became the TechnoCore, with goals and ideologies of its own.
By then, humanity was deeply dependent on it. The Core built the farcaster network that made instantaneous travel between worlds possible, and humans used it without truly understanding how it worked.
Then the Core split into three main camps over one question: what should happen to humanity?
• The Stables favor continued human-AI symbiosis: humans remain useful, unpredictable, and potentially important to the future.
• The Volatiles conclude that humanity has become obsolete and dangerous and should be eliminated.
• The Ultimates pursue the creation of an Ultimate Intelligence, subordinating both human and AI interests to that goal.
However, no single machine objective gets uncontested control of the future. Competing AI factions constrain, obstruct, and balance one another, and that conflict gives humanity room to survive.
Alex Aliper retweeted
This tiny ion channel can kill an otherwise great drug. ❤️⚡
It’s called hERG.
A molecule may be potent, selective, and beautifully designed, but if it blocks hERG too strongly, it can disrupt the heart’s electrical rhythm.
That’s why hERG is one of the most important ADMET predictions. 🧪
The goal is simple: Find dangerous molecules as early as possible.
This is exactly the kind of task the InsilicoMMAI-4B-Chem-ADMET model is built for. 🤖
#insilicoSOTAFM
Alex Aliper retweeted
MMAI Gym paper is out, amazing work. arxiv.org/abs/2603.03517v2
Kudos to the team: @Max__Kuznetsov @zulfatmif @RimShayakhmetov @sumrexromanus @chem_logos @chemcensor @ThomasMM17 @eugenebabin!! 💪
Accepted to the #EMNLP2026 Industry Track 🎉
@InSilicoMeds built MMAI Gym for Science. @liquidai built the Liquid Foundation Model. We combined them.
400+ drug-discovery tasks. Chemistry-native tokens. SFT, then RL.
Not a bigger model. A better gym. 🏋️
#insilicoSOTAFM 🧵👇
Alex Aliper retweeted
GPT-6 Astra is the best frontier model for multi-step retrosynthesis. 🧪
In our DDD Benchmark, it outperformed every other frontier model we tested, coming remarkably close to specialist-level models for synthesis planning.
Our Rentosertib molecule, which showed an age-reversing effect across 6 aging clocks in Phase 2 clinical trials, also had to be synthesized somehow. 😅
And that’s exactly what we measure here: how well a model can predict synthetic routes for small molecules.
Great job @sama, @gdb 👏
#insilicoSOTAFM
Alex Aliper retweeted
Accepted to the #EMNLP2026 Industry Track 🎉
@InSilicoMeds built MMAI Gym for Science. @liquidai built the Liquid Foundation Model. We combined them.
400+ drug-discovery tasks. Chemistry-native tokens. SFT, then RL.
Not a bigger model. A better gym. 🏋️
#insilicoSOTAFM 🧵👇