A true platform enables highly differentiated internal development. When this happens, a flywheel starts: others will line up to pay you for access, accelerating further R&D. It's an extremely high bar, but when it works, it's the best business model in biotech!

“Seemingly conscious AI” will come fast and permeate our lives Markets have always learned to exploit our evolutionary weak points Cigarettes, processed food, social media, etc exploit dopamine. Funeral homes, wedding planners, and baby gear retailers exploit other pathways

Nice complement to proteomics with Ebert and team here CRBN glues aren’t lucky hits anymore, establishing substrate rules for ZNFs Regulome proteins once out of reach now look more tractable eg. ZBTB11 in KRASi resistance and ZBTB7A/WIZ in HbF control

I had a suspicion that "extreme" general health advice based on studying bodybuilders and frail old people may have been problematic Nice, rational analysis and discussion here on what seems like a reasonable middle ground

I’m married and I have kids and I’m a lot more driven than I was in my 20s Kids are an incredible motivator

Biotech “failures” rarely come down to science alone Investor sentiment, strategic bets, regulatory curveballs, and commercial reality all loom large, esp. today Important analysis of where things break and what founders/boards can actually control.

part of my job at NOETIK is writing where, you ask? on a brand new substack: noetikai.substack.com. you should subscribe! why? because, tomorrow morning, we will release the first of three case studies on how our internal virtual-cell model has been applied to

Nice to see open science, and sharing data that didn't fit the initial hypothesis

Love to see this new technique finally out CRISPR scanning of chromatin accessibility A powerful way to move from correlation towards causation in the genome regulation, uncovering which bases truly control transcriptional networks Congrats to Brian Liau lab and team

McKnight’s work reminds us how deeply complex the regulome is So much biology is driven by unstructured, un-model-able domains invisible to AI Not only that, but he helped translate this to patients, delivering one of the only drugs for an 'undruggable' TF in the last 25 yrs

All genes have the same function: - Template gene products. The products are diverse. Their functions are complex and incompletely characterized.

Huge congrats to Michelle and the whole team! Real-life AI is already changing how work in the lab is done, and we can't wait to see what's next

Can we understand biology from "infinite" amounts of perturbation data ? - Only if we measure the relevant molecules. Perturbations and data scale 𝐜𝐚𝐧𝐧𝐨𝐭 compensate for missing the relevant molecules. 🧵

Alex Federation Lindsay K Pino - @[email protected] Will Fondrie and Gaelle Mercenne kicking off the 2025 Regulome Summit discussing how we use our regulome profiling platform and AI/ML here at Talus Bio to further cancer drug discovery.

The Regulome Summit opened this week with two local chordoma patients along with the Northwest Sarcoma Foundation, and Stephanie Schaub sharing their stories. A glimpse into the honest, sobering reality of living with an incurable regulome disease. Deeply inspiring, and set the

compiled case studies of how one can use a virtual cell that i posted to noetik.blog the cover photo is quite nice...small house next to big mountain. archetypically beautiful. colorado is a lot like this 5.5k words, 25 minute reading time owlposting.com/p/how-do-you-u…

Today in Science Magazine, we report a new DNA editing technology to seamlessly write massive changes into the right place in the human genome. The reason gene editing hasn't transformed human health is that current gene editing technologies like CRISPR are very limited. The

Baker Lab and Institute for Protein Design are on fire. AI-designed proteins to bind disordered TFs a few weeks ago, not we can bind the DNA itself too. A brand new toolkit for genome control is appearing before our eyes.

🚀🚀A new David Baker paper from Science Magazine is here👇 Can deep learning models finally overcome the massive computational challenge of predicting which 200 million possible human protein pairs actually interact in living cells?Institute for Protein Design UT Southwestern Medical Center "Predicting

We hear so much about the benefits of "antioxidants" without clear mechanism or nuance What if we could see how redox states reshape our biology? New tech like PACREDOX can map protein oxidation in 1000s of sites at once A window into how cells sense and respond to stress