Fay Lin

GEN Editors Reflect on Six of the Biggest Stories of the Year For the scientific enterprise, 2025 was marked by setbacks and challenges. But scientists are no strangers to adversity—the ability to overcome obstacles is built into our training. In turn, research ...

Boltz-2 is one of GEN News's six biggest stories of 2025! 📰Congrats to the Boltz-2 team and thank you @faylinphd.bsky.social for the recognition!
www.genengnews.com/industry-new... #AISky

The Challenge is sponsored by NVIDIA, @10xgenomics.bsky.social, and Ultima Genomics.

NeurIPS 2025: Altos Labs Wins Generalist Prize at Arc's Virtual Cell Challenge The inaugural competition surprisingly awarded two $100,000 grand prizes to address challenges in defining robust benchmarks for complex biology.

Altos Labs has taken home the Generalist Prize for the most "well-rounded" model across metrics at @arcinstitute.org's Virtual Cell Challenge!

The competition highlights the challenge of defining robust benchmarks for biology.

#NeurIPS2025

Read at GEN:
www.genengnews.com/topics/artif...

AI in Protein Design: Hype vs. Reality Explained by David Baker In this GEN interview, Nobel Laureate David Baker, PhD, emphasizes that designing proteins from scratch is reality and unpacks what's needed for AI to transform medicine.

AI in Protein Design: Hype vs. Reality Explained by David Baker
In this GEN interview, he emphasizes that designing new proteins from scratch is now a reality. Whether AI transforms medicine will require improving our understanding of biology's complexity

The State of AI in Drug Discovery 2025, Wed, Oct 29th, 2025 @ 11:00 AM Mary Ann Liebert, Inc. publishers Artificial intelligence (AI) and machine learning partnered with big data generation at unprecedented scale is transforming the worlds of biotechnology and pharma, particularly in the arena of drug discovery. From generative AI to unlock novel drug candidates to virtual cells that gleam insights across multimodal biology, we are witnessing the cusp of an exponential curve of AI innovation that is poised to enhance and perhaps overhaul the design and validation of novel therapeutics. In 2024, GEN debuted our virtual summit, The State of AI in Drug Discovery, which drew a record audience. This year’s summit provides a platform for an outstanding group of researchers, evangelists and entrepreneurs to showcase the latest advances, applications and challenges regarding the deployment of AI across therapeutic development. The program features many of the most exciting companies and organizations powering the AI revolution, including NVIDIA, Xaira Therapeutics, Recursion, Chan Zuckerberg Initiative, Arc Institute, Generate: Biomedicines, and more! Please join us on October 29th for a dynamic program featuring: A featured presentation with Bo Wang, PhD (Xaira Therapeutics) on the promise of multimodal AI for health care The democratization of small molecule drug discovery with the developers of Boltz-2, Gabriele Corso, PhD (MIT), Regina Barzilay, PhD (MIT), and Najat Khan, PhD (Recursion) Evolving industry trends across the AI therapeutic landscape with Derek Lowe, PhD (Novartis and author of "In the Pipeline"), Molly Gibson, PhD, (Flagship Pioneering), and Stacie Calad-Thomson, PhD (NVIDIA) Building the virtual cell with Theofanis Karaletsos (Chan Zuckerberg Initiative), Emma Lundberg, PhD (Stanford), Hani Goodarzi, PhD (Arc Institute), and Ron Alfa, MD, PhD (Noetik) What’s next for AI in protein design with Surge Biswas, PhD (Nabla Bio), Tharindi Hapuarachchi, PhD (Generate: Biomedicines), and Oliver Vince, PhD (Basecamp Research) Breakout sessions from our event sponsors, Revvity Signals, Benchling, and Telesis Bio. Registration for The State of AI in Drug Discovery is entirely free. Sponsored by

Fantastic free virtual summit 11 am
ET today:
The State of AI in Drug Discovery

Hosted by @faylinphd.bsky.social
Speakers include @dereklowe.bsky.social
@emmalundberg.bsky.social
@bowang87.bsky.social
@mollygibson.bsky.social
Regina Barzilay
… and more!

webinars.liebertpub.com/e/The-State-...

The work was led by @adambroerman.bsky.social at @uwproteindesign.bsky.social!

AI-Designed Protein Switches Control Drugs with Speed David Baker’s Nobel Prize-winning lab has designed protein on/off switches as a tool for tunable cancer immunotherapies and biosensors.

David Baker's Nobel Prize-winning lab has now designed a molecular on/off switch to control the speed of protein binding, supporting safer medicines that quell harmful side effects and activate drugs on cue, and sensitive biosensors for SARS-CoV-2 testing!

www.genengnews.com/topics/artif...

AI Designs Viable Bacteriophage Genomes, Combats Antibiotic Resistance AI-guided design of 16 functional bacteriophage genomes offers a path for phage-based therapies against antibiotic-resistant infections.

Genome foundation models, Evo 1 and Evo 2, have now generated viable bacteriophage genomes, demonstrating experimental validation of whole genomes designed by AI!

@arcinstitute.org @brianhie.bsky.social @samuelhking.bsky.social

Read more at GEN:
www.genengnews.com/topics/artif...

@martinpacesa.bsky.social @sokrypton.org @pranam.bsky.social

BindCraft AI Model Achieves One-Shot Functional Protein Design The open-source pipeline designs proteins with improved success rates for therapeutic targets, including gene editing proteins and allergens.

BindCraft is now published in @nature.com!

The open-source pipeline for de novo protein binder design achieves significantly improved success rates with nanomolar affinity and therapeutic potential from gene editing to allergens.

Read more at GEN:
www.genengnews.com/topics/artif...

Given the high uptake for a “Cas9 look-alike,” the team is inspired to expand toward new territories, including base editors, prime editors, large insertion techniques, and more to demonstrate the generalizability of its protein design platform. 4/4

Since OpenCRISPR-1's open source release in April 2024, tens of thousands of academic and industry researchers have accessed the sequence across a variety of verticals, from developing drought-resistant crops to drug discovery. 3/

Despite being hundreds of mutations away from any known natural protein. Profluent's LLM generated CRISPR proteins show improved activity, specificity, and immunogenicity relative to naturally occurring gene editors. 2/

Profluent’s AI-Designed Gene Editor Provides Glimpse Into Generalizable Platform OpenCRISPR-1, an AI-generated custom gene editor capable of precision editing of the human genome. is just the beginning of Profluent’s AI applications.

OpenCRISPR-1, Profluent's AI-designed gene editor, is now published in Nature!

@thisismadani.bsky.social emphasizes the need to move away from random discovery and into bespoke design using AI. 1/

Read more at GEN:
www.genengnews.com/topics/artif...

Jura Bio’s AI Data Loop Enables Large-Scale De Novo Antibody Design VISTA, an AI-controlled data loop system, guides large-scale antibody design from scratch for applications in cancer therapy.

Great coverage of VISTA by the wonderful @faylinphd.bsky.social over at GEN: www.genengnews.com/topics/artif...

Read about VISTA: www.jurabio.com/blog/vista

Undruggable No More: AI Hits Disordered Proteins, Unlocks Therapy Targets David Baker’s lab has successfully designed binders to disordered proteins, expanding therapeutic access to over 50% of the human proteome.

David Baker's Nobel Prize-winning lab has now designed binders to "undruggable" disordered proteins, unlocking therapeutic access to over 50% of the human proteome!

‪@uwproteindesign.bsky.social‬

Read more at GEN:
tinyurl.com/4fw6wvkm

Democratizing Artificial Intelligence in Pre-Clinical Drug Discovery While AI-driven approaches tout increased speed and lower costs, commercial interests compromise scientific collaboration.

Excited to see our new AI Small Molecule Drug Discovery Center #MountSinai featured in GEN!
🧬🧠 Read more: genengnews.com/topics/artif... Thank you @faylinphd.bsky.social
#AI #DrugDiscovery #Genetics

As a baseline, competitors will initially go head-to-head with Arc’s first virtual cell model, STATE, which is designed to predict how various stem cells, cancer cells, and immune cells respond to drugs, cytokines, or genetic perturbations. 3/3

The initiative is a step toward virtual cell benchmarking and acceleration, and follows in the footsteps of the CASP competition, the biannual experiment that assesses the latest state-of-the-art models in protein structure prediction, which enabled Nobel Prize-winning, AlphaFold. 2/

Arc Institute Launches Virtual Cell Challenge to Accelerate AI Model Development The open benchmark competition will evaluate the ability of AI-powered virtual cell models to generalize to new cell contexts for therapeutic applications.

@arcinstitute.org‬ has announced the inaugural “Virtual Cell Challenge,” sponsored by NVIDIA, @10xgenomics.bsky.social and Ultima Genomics,
which will evaluate the ability of AI models to generalize to new cell contexts for therapeutic applications. 1/

www.genengnews.com/topics/artif...

Xaira Therapeutics Releases Largest Perturb-Seq Dataset to Power the Virtual Cell The AI unicorn's Perturb-seq atlas is publicly available and detects dose-dependent genetic effects to enhance predictive power for drug discovery.

Xaira Therapeutics has released the largest publicly available Perturb-seq dataset to train the AI-powered virtual cell for drug discovery!

The work is described in Xaira's first preprint, capping a year after the company's $1 billion launch.

www.genengnews.com/topics/artif...

Million Species Listing: Basecamp Research Unearths Trove of Sequence Data From Novel Species BaseData aims to improve AI models in biology by offering a 10-fold expansion of protein diversity compared to all public databases combined.

More in GEN from @faylinphd.bsky.social: www.genengnews.com/topics/artif...

In collaboration with @recursionpharma.bsky.social, the work was led by Gabriele Corso @gcorso.bsky.social, Saro Passaro, and Jeremy Wohlwend from the lab of Regina Barzilay at @mit.edu!

Boltz-2 Released to Democratize AI Molecular Modeling for Drug Discovery The MIT model predicts molecular binding affinity at newfound speed and accuracy, offering a powerful tool for commercial drug discovery.

Boltz-2 now predicts molecular binding affinity at newfound speed and accuracy - and it's open source!

The permissive MIT license allows commercial drug developers to apply proprietary data and offers an advance over the structure prediction capabilities of AlphaFold 3.

tinyurl.com/37c96c6y

Scratch That? De Novo Antibody Design Enters the AI Drug Discovery Toolbox Debates over AI antibody design terminology have clouded the industry’s shared mission of bringing better therapeutics to the clinic faster.

AI hype or reality - can we design antibodies from scratch (de novo)?

In my latest for GEN, I unpack where de novo design is today and how arguments over terminology have clouded the industry’s shared mission of bringing better therapeutics to the clinic faster.

tinyurl.com/4uamc8c5

Excellent story from @faylinphd.bsky.social on the hot news from #ASGCT2025 — @kiranmusunuru.bsky.social and colleagues tailor the world’s first bespoke gene editor to treat KJ, an infant with an ultra-rare genetic disease, heralding a new era in precision med.
www.genengnews.com/topics/genom...

Imprint Labs Launches to Pioneer "Forensic Immunology" For Chronic Disease Imprint has raised $15 million to decode the body's immune memory for understanding chronic diseases, such as long-COVID and psychiatric disorders.

📣Major thanks to @faylinphd.bsky.social at GEN News for this lovely piece on IMPRINT! It was a pleasure for our founders — @beckbrachman.bsky.social, @victorgreiff.bsky.social, and Raymond Alvarez — to speak about our mission and research, and being NYC's first FRO.

@convergentresearch.bsky.social

ASGCT 2025: World’s First Patient Treated with Personalized CRISPR Therapy Treated with an individualized gene-editing therapy that corrects mutations directly on the genome, a patient shows improvement from a rare metabolic disease.

BREAKING NEWS: World’s first patient treated with personalized CRISPR therapy!

Born with a severe metabolic disease with a 50% mortality rate, an infant has improved after receiving an individualized gene editing therapy, a glimpse of a new era in genomic precision medicine.

tinyurl.com/m52udxht

A "ChatGPT prompter" for interrogating single-cell transcriptomics, TranscriptFormer probes cellular biology across organisms, with predictive capabilities for disease state classification and tissue context, and generative tasks, such as inferring cell type-specific transcription factors. 2/2

The Cross-Species Cell Atlas: Chan Zuckerberg Initiative Releases TranscriptFormer AI Model Trained on data covering 1.5 billion years of evolution, CZI’s new generative AI model probes cellular biology across species for therapeutic applications.

Introducing TranscriptFormer, the cross-species cell atlas!

This new generative #AI model from @chanzuckerberg.bsky.social is trained on data from over 110 million cells from 12 different species, covering 1.5 billion years of evolution.

tinyurl.com/2yet8k42