Ghazaleh Madani’s Post

If you walked into a biotech lab five years ago and then walked into one today... You'd think you entered a completely different industry. The biggest transformation isn't happening in the science. It's happening in how we do the science. A few years ago, biotech was mostly about generating data. Today, the challenge isn't generating data. It's generating the right data, faster than ever before. Over the last few years, I've had the opportunity to work in different parts of biotechnology—from CAR-T research in academia to high-throughput sequencing and now biofoundry-driven genomics. One thing has become very clear. The industry is changing faster than many people realize. What once took weeks can now happen in days. What took days is now measured in hours. And in some cases, automation is making it possible to generate results the very same day. Robotics, automated liquid handlers, high-throughput sequencing, and smarter laboratory workflows are no longer "nice to have." They're becoming the foundation of modern biotechnology. But what excites me most isn't that machines are replacing scientists. They're removing repetitive work so scientists can spend more time solving problems, designing better experiments, and asking bigger questions. I believe the next wave of innovation won't come from a single breakthrough technology. It will come from how seamlessly biology, automation, artificial intelligence, and data science work together. We're moving toward laboratories that are faster, more reproducible, more scalable, and capable of accelerating discoveries that once seemed years away. There's still plenty of room to improve—better interoperability between instruments, smarter automation, more accessible genomic technologies, and AI systems that don't just analyze data but actively help design experiments. It's an exciting time to be in biotech. Not because we've reached the future. But because we're building it every day. Curious to hear what others in biotech think. What's the biggest change you've seen in our industry over the last few years, and where do you think we're headed next? #Biotechnology #Genomics #NGS #SyntheticBiology #Automation #Innovation #FutureOfBiotech

CuspAI Is Using AI to Speed Up Scientific Discovery! Some of the world's biggest technological breakthroughs begin with a new material. Better batteries, faster semiconductors, cleaner industrial processes, and more effective carbon capture all depend on discoveries that can take years or even decades to achieve. AI has the potential to dramatically shorten that timeline by helping researchers identify the most promising candidates before they ever enter the lab. CuspAI is building AI models that predict how new molecules and materials will behave, allowing scientists to focus their experiments where they're most likely to succeed. Backed by $450 million in Series B funding, the company is betting that the next wave of innovation won't just come from writing better software, but from discovering better materials. https://lnkd.in/gbQNRjKS CuspAI Ben Amy Chad James Anand Willis Alessandro Lawashan Vincent Vikram Zheela Ruth Nayoung Debbie Will Markus Max Yvonne Gloria Austin Tin Mark Samir Thomas Felix Joe Nicholas Elena Govert Jonas Hannah Ry Jetze Aron Friso Floor Nathalie Marion Anita #ArtificialIntelligence #MaterialsScience #DeepTech #ScientificDiscovery #MachineLearning #ClimateTech #Innovation #FutureTech #Research #Technology

🚨The U.S. Department of Energy (DOE) just announced $10 million in non-dilutive funding for small businesses working in AI, biotech, quantum, and materials science. It's part of something bigger: the White House put more than $5 billion behind the Genesis Mission this week, and this $10 million SBIR/STTR tranche is DOE's way of pulling small companies directly into it. The Genesis Mission is the Trump Administration's new national push to embed AI directly into the scientific process, with a stated goal of doubling the productivity of American research within a decade. It spans 15+ different agencies, including DOE, HHS, NSF, and NIST, and runs through the national labs (DOE's Berkeley Lab alone is leading 13 projects under it). It's aimed at four areas: scaling biotechnology, building working quantum computers, designing new materials, and creating AI-driven "self-driving" labs that run their own experiments. This SBIR/STTR Phase I opportunity is the on-ramp for small companies to get funded work inside that mission. DOE expects to make roughly 40 Phase I awards across those same four topic areas: biotechnology, quantum computing and networking, advanced materials, and AI-driven autonomous laboratories. And this is just the opener: a broader DOE Phase I opportunity covering additional technology topics is coming later this fall. DOE wants a short Pitch application first, due Thursday, September 10, and only the companies it invites back after that review get to submit a full application. Everything runs through the new DOE SBIR/STTR Application Hub. Link to the Phase I application and full announcement in the comments below. #AmTech #AI #SBIR #Biotech #Quantum

A terrific analysis and thoughtful critique by Rob Miller, especially about the flow from science to engineering, to impact and seeing the reverse flow too... with designers and engineers tackling wicked challenges, thereby stimulating new scientific enquiry, and posing new questions for science, with AI accelerating those flows. The UK is home to world leading Science, Engineering and Design Institutions and extraordinary talent. Also, our ecosystem of start ups and scale ups leads in Europe. This is an important read that can help shape not only policy but collaboration between institutions, industry and investors.

Rob Miller of the University of Cambridge’s Whittle Laboratory makes a hugely important point here: the US has multi-layered and, crucially, flexible systems. Prof. Miller’s closing point is well made - in my experience, winning Government approval is a matter of seeking forgiveness, not permission; creating strong forward momentum and building the structures we want to see, even if only at a sectoral and/or regional level. That means set-reliance and, yes, self-funding. When those structures can credibly evidence their ability to scale up and help deliver the Government of the day’s objectives, THAT’s what commands Government attention. Many a regional initiative has floundered precisely because its leaders wanted ministerial benediction before actually **doing** anything - and so could not persuade officials and departments to act. Prof. Miller is right to lead the charge and Eastern Promise stands ready to support this. Productivity East Stuart Catchpole James Allen Scott Cogman Joanne Kerrison Joseph Corrigan Uday Phadke Will Bridgman Katy Davies Charlotte Horobin Al Kingsley MBE

The deeper the tech, the more likely it started in a lab. That's the clear pattern in new data from the Spotlight on Spinouts report, produced with the Royal Academy of Engineering Enterprise Hub. Overall 23% of all VC-backed deep tech startups in the UK are university spinouts, but there is also large variance within different sectors. Looking at UK VC-backed deep tech startups since 2010: 🔦 58% of photonics startups are academic spinouts ⚛️ 56% of quantum 🩺 48% of medical devices At the other end: AI (17%), robotics (14%), space (8%). The more science- rather than engineering-intensive the technology, the more likely it began as research inside a university in the UK. Understanding the startup and spinout landscape also highlights real implications for backing frontier innovation. If the UK wants to lead in quantum, photonics, and next-gen medicine, the pipeline runs straight through universities. This makes tech transfer, spinout terms, and early-stage capital a national competitiveness question, not just an academic one.

🚀 The 2026 EIC Pathfinder Challenges are now open! A new funding opportunity is available for visionary, high-risk, high-impact research projects with the potential to create breakthrough technologies and shape the future of Europe. The 2026 EIC Pathfinder Challenges focus on tackling ambitious scientific and technological challenges that require interdisciplinary collaboration and bold ideas. These year's challenges are: ⚡ Advanced Materials for Miniaturised Energy Harvesting Systems 🧬 Biotechnology for Healthy Ageing 🤖 DeepRAP: Deep Reasoning, Abstraction & Planning towards trustworthy Cognitive AI Systems If you're part of a research institution, university, startup, SME, or innovation ecosystem working on transformative solutions, this is an excellent opportunity to explore. Learn more here: https://lnkd.in/dd-5HNHG #EIC #EICPathfinder #Challenges

I spoke with Mehdi Farza, PhD. about the new Rowan–Vicena integration, how I got into this field, what we're building at Rowan, and why I'm excited about integrating simulation-based toolkits with AI agents. Rowan's chemistry and materials-science workflows are now available in Vicena: https://lnkd.in/gV8ABWn2

This saves $50K just in license cost per year! 36 Workflows in Computational Chemistry and Molecular Dynamics is now supported in Vicena 3.5. We charge the cost price of 4 cents per CPU minute! Maybe crazy but we want to make sure small startups scientists are getting access to best tools at the lowest price. This is the interview with the co-founder at Rowan, Corin Wagen. I am very proud of this collaboration! What can it be done with it: 1. Dock a molecule into a protein and see if it fits: You have 50 ideas for molecules to make. You dock all 50 into the crystal structure, look at the poses, and eliminate the ones that don’t fit the pocket or clash with key residues. 2. Run ADMET + solubility + pKa on a list of molecules before you make them: ou have 200 compounds from a virtual screen. You run this before ordering synthesis. Half of them have a hERG alert or are predicted to be insoluble at pH 7.4. You drop them. 3. Predict relative binding affinities with FEP on a congeneric series: You have 10 analogs and a finite budget. FEP tells you which 3 to actually make and assay first. If the predictions are good (and they are, within ~1 kcal/mol for well-behaved congeneric series) Who is this for? 1. Small molecule drug discovery startups, 2. Academic drug discovery labs and core facilities, 4. Agrichemical and materials chemistry startups. We have limited seats available so please write me if you want to work with it.

Predicting how tightly a molecule binds to its target is one of the fundamental challenges in drug discovery. Until now, researchers have largely faced a tradeoff: fast methods sacrifice accuracy, while more accurate approaches are often too computationally expensive to use at the scale modern drug discovery demands. Last year, our team at Recursion worked with the Massachusetts Institute of Technology to advance the field with Boltz-2, showing that AI could make binding affinity prediction both more accurate and more accessible. Today, we are excited to share the next release from our Valence Labs: Nesso-1, which we are also open sourcing. For years, there’s been an assumption that achieving better binding affinity prediction requires increasingly detailed, atomistic models. But Nesso-1 builds on thoughtfully-designed course-grained representations architecture to achieve comparable predictive performance at a fraction of the computational cost. The result is a model that predicts binding affinity in approximately 1 second on a single GPU – up to 20× faster than Boltz-2 – while matching or surpassing its performance across public benchmarks and, importantly, a diverse set of internal biochemical assays that better reflect real-world medicinal chemistry. By leveraging NVIDIA cuEquivariance, we've also accelerated training and inference by an additional 2–3×. Why does this matter? Because in AI-native drug discovery, speed isn't just an engineering metric – it determines how much science you can explore. Faster affinity prediction means screening much larger chemical libraries, running many more design iterations, and bringing autonomous drug discovery workflows closer to reality. Instead of forcing researchers to choose between computational efficiency and predictive performance, it demonstrates that thoughtful model design can advance both simultaneously. We're already using Nesso-1 internally alongside our proprietary datasets to accelerate active drug discovery programs, and we're excited to continue improving it with the community. Open science accelerates innovation. We hope Nesso-1 becomes a foundation the community can build on as we collectively push toward faster, more effective drug discovery. Weights and code are fully open source: GitHub: https://lnkd.in/grwpVc5Z HuggingFace: https://lnkd.in/g5ST7scJ Technical Report: https://lnkd.in/gyTgU7u3 #Nesso1 #DrugDiscovery #ArtificialIntelligence #OpenScience #MachineLearning #TechBio