Lightnews — Scholar-powered news

Arc Institute @arcinstitute.org · 18h

Arc's community continues to grow this fall with our 9th Core Investigator John Pluvinage and 3rd Science Fellow @mayaarce.bsky.social joining us in Palo Alto, along with 7 Innovation Investigators and 15 Ignite Awardees at our partner universities.

Learn more:
arcinstitute.org/news/faculty...

Arc Institute Adds Two New Labs; Announces 2025 Cohort of Innovation Investigators and Ignite Awardees | Arc Institute

We are excited to launch three parallel searches for our curiosity-driven Laboratories.

arcinstitute.org

1 2

Arc Institute @arcinstitute.org · 12d

The team is already working to expand the platform’s abilities, including testing in clinically relevant immune and stem cells, and engineering future versions of the system that can rearrange sequences beyond one megabase.

Learn more in the full paper: www.science.org/doi/10.1126/...

Megabase-scale human genome rearrangement with programmable bridge recombinases

Bridge recombinases are naturally occurring RNA-guided DNA recombinases that we previously demonstrated can programmably insert, excise, and invert DNA in vitro and in Escherichia coli. In this study,...

www.science.org

2 4

Arc Institute @arcinstitute.org · 12d

The optimized ISCro4 system was then tested in proof-of-concept edits – including excision of Friedreich’s ataxia–associated repeats and deletion of BCL11A , a validated sickle cell target – suggesting the technology's potential for therapeutic application.

1 2

Arc Institute @arcinstitute.org · 12d

Through systematic engineering of both the recombinase protein and its guide RNA guide, the researchers optimized ISCro4, reaching up to 20% insertion efficiency and 82% specificity for hitting its intended targets in the human genome.

1 1 4

Arc Institute @arcinstitute.org · 12d

Building on previous work in @nature.com that introduced bridge recombinases in bacterial systems, the research team tested 72 natural candidates in human cells.

While few had measurable activity, one system, ISCro4, was suitable for further optimization.

1 2

Arc Institute @arcinstitute.org · 12d

For decades, human genome editing has been limited to small, localized modifications.

Today, in a new paper published in @science.org , researchers from Arc's Hsu lab show that bridge recombinase technology is capable of large-scale genomic rearrangements in human cells.

2 14 23

Arc Institute @arcinstitute.org · 13d

New work from Arc's @pauldatlinger.bsky.social of our Genome Engineering Technology Center. Congrats!

pauldatlinger.bsky.social @pauldatlinger.bsky.social · 13d

CAR T cells showcase the enormous potential of cell therapies, but often fail due to lack of evolutionary optimization. Today in @nature.com , we use #CELLFIE to engineer cell therapies at scale and share the largest resource of CRISPR screens in CAR T cells. www.nature.com/articles/s41...

Systematic discovery of CRISPR-boosted CAR T cell immunotherapies - Nature

CELLFIE, a CRISPR platform for optimizing cell-based immunotherapies, identifies gene knockouts that enhance CAR T cell efficacy using in vitro and in vivo screens.

www.nature.com

2 3

Reposted by Arc Institute

Xiaojing Gao @synbiogaolab.bsky.social · 13d

Having often dealt with the frustration of binder-limited projects, we sought a more accessible source for nanobodies than yeast display or llama. Here we introduce Germinal, computationally designing antibody-like binders with such a hit rate that only tens need to be screened for each target.

1 3 7

Arc Institute @arcinstitute.org · 13d

Or go behind-the-scenes with the co-authors in this conversation about the work: arcinstitute.org/news/buildin...

Building “Germinal” for AI-Designed Antibody Molecules | Arc Institute

Going from designing individual genes to complete genomes is an incredibly challenging problem. We have previously shown that the genomic foundation models like the Evo series can generate single prot...

arcinstitute.org

2

Arc Institute @arcinstitute.org · 13d

With plans to make Germinal freely available, the team is lowering the barrier for antibody discovery and opening new doors for both basic research and potential therapies.

Learn more in the full preprint: www.biorxiv.org/content/10.1...

Efficient generation of epitope-targeted de novo antibodies with Germinal

Obtaining novel antibodies against specific protein targets is a widely important yet experimentally laborious process. Meanwhile, computational methods for antibody design have been limited by low su...

www.biorxiv.org

1 2 2

Arc Institute @arcinstitute.org · 13d

When tested on four unique targets, the team found Germinal could deliver working nanobodies with nanomolar affinities.

From just tens of designs per target, they achieved success rates of up to 22%, a significant jump from existing approaches that often require thousands.

1 2

Arc Institute @arcinstitute.org · 13d

To test candidates, the team created a two-stage pipeline.

A split-luciferase assay rapidly tests expression and binding directly in cell culture, allowing efficient filtering. Only the most promising candidates advance to more detailed biophysical characterization.

1 2

Arc Institute @arcinstitute.org · 13d

The team introduced custom rules to ensure binding occurs primarily through CDRs rather than framework regions, loops remain flexible instead of rigid, and sequences reflect natural antibody-like features.

The result is computational designs that behave like true antibodies.

1 2

Arc Institute @arcinstitute.org · 13d

Germinal works by integrating AlphaFold-Multimer, which predicts protein–protein structures, with IgLM, an antibody-specific language model.

By jointly optimizing structure and sequence, it redesigns CDR loops on a fixed antibody framework to target chosen epitopes.

1 2

Arc Institute @arcinstitute.org · 13d

In another preprint from the @brianhie.bsky.social Lab and @synbiogaolab.bsky.social, they introduce Germinal, a generative AI system for de novo antibody design.

Germinal produces functional nanobodies in just dozens of tests, making custom antibody design more accessible than ever before.

1 14 16

Arc Institute @arcinstitute.org · 19d

Bao also recently appeared on the Nature Biotechnology podcast to discuss ways researchers are helping electronics interface with the body. www.nature.com/articles/s41...

Forum: Bao and Rogers - Nature Biotechnology

Nature Biotechnology - Forum: Bao and Rogers

www.nature.com

2

Arc Institute @arcinstitute.org · 19d

“We hope to thread these thin electronics inside and throughout organoids, to promote and monitor their growth,” Bao told the Stanford Report.
news.stanford.edu/stories/2025...

Soft bioelectronic fiber can track hundreds of biological events simultaneously

Developed by Stanford researchers, NeuroString is a hair-thin multichannel biosensor and stimulator with promising potential applications in drug delivery, nerve stimulation, smart fabrics, and more.

news.stanford.edu

1 1

Arc Institute @arcinstitute.org · 19d

Published in @nature.com, the work offers a powerful platform for minimally invasive implantable electronics, where diverse sensing and stimulation functionalities can be effectively integrated. www.nature.com/articles/s41...

High-density soft bioelectronic fibres for multimodal sensing and stimulation - Nature

High-density multimodal soft bioelectronic fibres provide a platform for minimally invasive implantable electronics, where diverse sensing and stimulation functionalities can be effectively integrated...

www.nature.com

1 1

Arc Institute @arcinstitute.org · 19d

Arc Innovation Investigator research highlight:

Stanford Professor Zhenan Bao and colleagues have developed NeuroString, a hair-thin multichannel biosensor and stimulator with promising potential applications in drug delivery, nerve stimulation, smart fabrics, and more.

1 1 3

Reposted by Arc Institute

Fay Lin @faylinphd.bsky.social · 20d

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...

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.

www.genengnews.com

5 8

Arc Institute @arcinstitute.org · 20d

Learn more from @brianhie.bsky.social and first author Samuel King about how they built the first AI-generated genomes: arcinstitute.org/news/hie-kin...

How We Built the First AI-Generated Genomes | Arc Institute

Going from designing individual genes to complete genomes is an incredibly challenging problem. We have previously shown that the genomic foundation models like the Evo series can generate single prot...

arcinstitute.org

6

Arc Institute @arcinstitute.org · 20d

Nearly 300 promising candidates were chemically synthesized, assembled, and introduced into E. coli cells to test their viability.

Compared to ΦX174, the 16 viable designs carried hundreds of mutations, yet retained all essential functions needed for replication and infection.

1 6

Arc Institute @arcinstitute.org · 20d

They then fine-tuned Evo 1 & Evo 2 on ~15,000 Microviridae genomes.

This allowed the models to learn the specific “grammar” and patterns of ΦX174 and generate thousands of new sequences, each proposing novel combinations of the 11 genes required for phage function.

1 5

Arc Institute @arcinstitute.org · 20d

As a test case, the team focused on E. coli C and its native phage ΦX174.

This system is safe, well-characterized, and small enough to be tractable, yet still far more complex than anything AI had previously been asked to design at the genome scale.

1 5

Arc Institute @arcinstitute.org · 20d

They first evaluated whether Evo 1 & Evo 2 could generate phage-like sequences at all.

Using geNomad, BLAST, and protein structure predictions, they confirmed that the models are capable of generating realistic sequences with diversity beyond natural evolution.

1 6