## Can Nerai Bioscience's AI Platform Unlock the 80% of Pathogenic Mutations CRISPR Currently Misses?

University of Zurich spinout Nerai Bioscience launched in 2025 with a CHF 500,000 seed round from Kickfund, aiming to solve one of therapeutic [CRISPR-Cas9](https://synbiointel.com/glossary/crispr-cas9)'s most persistent engineering constraints: PAM sequence restriction. The company's platform combines automated [directed evolution](https://synbiointel.com/glossary/directed-evolution) running 100 parallel experiments simultaneously, high-throughput cellular screening across thousands of DNA target sites, and a machine learning feedback loop — all designed to generate customized Cas9 variants at a scale no manual engineering workflow can match. The immediate clinical rationale is stark: existing CRISPR tools can currently repair only approximately 20% of pathogenic mutations, according to the company. With roughly one in ten people worldwide carrying a genetic disease and more than 90% of those conditions lacking approved therapies, the addressable gap is enormous. Nerai has already built 8 disease-specific pipelines, with three — NB-301 (type I citrullinemia), NB-20X (phenylketonuria), and NB-101 (alpha-1 antitrypsin deficiency) — in active in vivo lead optimization.

---

## The Core Technical Thesis: PAM Flexibility at Industrial Throughput

The fundamental bottleneck Nerai is attacking is not editing efficiency within a targetable site — it is the fraction of the genome that is targetable at all. [CRISPR-Cas9](https://synbiointel.com/glossary/crispr-cas9) requires a protospacer adjacent motif (PAM) sequence adjacent to any cut site; different Cas9 orthologues and engineered variants recognize different PAMs, and the coverage of any single tool is therefore finite. Most clinical programs today are built around a handful of well-characterized Cas9 variants whose PAM profiles are well understood but narrow.

Nerai's platform addresses this by running what the company describes as 100 sets of parallel directed evolution experiments on multiple CRISPR enzymes simultaneously. The output is a continuously expanding library of novel enzyme variants with different PAM recognition capabilities. This is then filtered through high-throughput cellular screening: candidate tools are introduced into human cells and their targeted editing activity, accuracy, and editing efficiency are measured across thousands of DNA target sites. The machine learning layer then analyzes the resulting high-quality dataset — learning relationships between protein sequences, DNA target features, and actual editing outcomes — and uses those models to guide the design of the next evolution round.

The feedback loop is architecturally similar to what companies like [Cradle](https://synbiointel.com/companies/cradle) and [Absci Corporation](https://synbiointel.com/companies/absci) have built for antibody and general protein design, but Nerai's application is specifically optimized for the PAM-specificity and editing-window parameters that matter for therapeutic gene correction. Whether that domain specificity translates into a meaningful competitive moat — versus a larger, better-funded platform adapting to the same problem — is the question investors will need to answer in the next financing round.

---

## Pipeline Snapshot: Liver First, Then Ophthalmology

Nerai's 8 current pipelines reflect a deliberate disease-area focus rather than a spray-and-pray approach:

**Inherited liver diseases (in vivo lead optimization):**
- **NB-301** — Type I citrullinemia
- **NB-20X** — Phenylketonuria
- **NB-101** — Alpha-1 antitrypsin deficiency

**Inherited liver diseases (discovery stage):**
- **NB-401** — Homocystinuria
- **NB-501** — Classic type I galactosemia
- **NB-601** — Primary hyperoxaluria
- **NB-701** — Undisclosed indication

**Inherited ophthalmic diseases:**
- **NB-e101** — Autosomal dominant retinitis pigmentosa

Each pipeline uses customized PAM-specific Cas9 variants — not a single tool applied broadly. This is both the company's strength and its near-term execution risk. Developing, screening, and clinically validating a distinct engineered Cas9 for each indication is precisely the expensive, slow process the company claims to compress. The platform's value proposition only holds if the AI-guided evolution cycles demonstrably reduce preclinical timelines and per-program costs versus conventional approaches. At seed stage, that remains to be demonstrated with published data.

The liver disease focus is strategically sensible: the liver is well-served by lipid nanoparticle delivery, reducing the need to solve delivery alongside editing. Retinitis pigmentosa adds optionality in ophthalmology, another tissue where local delivery is relatively tractable.

---

## Team Credentials and Institutional Backing

The founding team's ETH Zurich and University of Zurich pedigree is credible for this technical domain. CEO Vincent Forster holds a doctorate in pharmaceutical sciences from ETH Zurich and previously co-founded Versantis, a liver disease biotech — directly relevant experience. Kim Fabiano Marquart (PhD, ETH Zurich) conducted CRISPR protein engineering and therapeutic gene editing research at the University of Zurich's Institute of Pharmacology and Toxicology. Lukas Schmidheini (PhD, ETH Zurich) leads the engineering platform. Sasha Melkonyan, with a master's in computational biology from ETH Zurich, heads machine learning.

The institutional lineage matters here. The University of Zurich and ETH Zurich ecosystem has produced credible biotech spinouts, and the team's intersection of wet-lab CRISPR engineering and computational biology is the right combination for the stated approach. Whether four co-founders at seed stage have sufficient bandwidth to run 100-parallel evolution experiments, human cell screening at scale, and ML model training simultaneously is a legitimate operational question.

Kickfund's CHF 500,000 seed is modest relative to what the platform buildout will require. The funding is explicitly earmarked for advancing preclinical development of the high-throughput technology and the rare disease pipelines — a reasonable use of seed capital, but the runway to Series A data will be tight.

---

## Industry Trajectory: What Nerai's Emergence Signals

The PAM-flexibility problem is not new — SpCas9's NGG PAM requirement was identified as a coverage limitation within years of the original editing demonstrations. The field has produced SpRY (near-PAMless), SaCas9, and numerous base editor variants to address subsets of the problem. What is changing is the throughput of protein engineering itself. Automated directed evolution combined with ML-guided design is compressing variant discovery cycles that once took years into months.

Nerai's positioning reflects a broader thesis: that the bottleneck in therapeutic gene editing is no longer understanding which mutations cause disease (genomics has largely solved that), nor even delivery (still hard, but tractable for liver and eye), but the availability of sufficiently precise, sufficiently active editing tools for the specific mutation each patient carries. If rare disease economics ever support truly personalized CRISPR tools — a high bar given clinical development costs — platforms that can generate and validate novel Cas9 variants rapidly become infrastructure.

The skeptical read: the rare disease gene editing space already has well-capitalized players, and the jump from a high-throughput screening platform to clinical-grade, GMP-manufactured, regulatory-approved therapies is long. CHF 500K funds experiments, not IND filings.

---

## Key Takeaways

- Nerai Bioscience closed a **CHF 500,000 seed round** (Kickfund) in May 2025 to advance AI-driven CRISPR protein engineering
- The platform runs **100 parallel directed evolution experiments** simultaneously, paired with high-throughput cellular screening across thousands of DNA target sites
- Existing CRISPR tools can correct only approximately **20% of pathogenic mutations** — Nerai's PAM-variant library approach targets the uncovered 80%
- **8 active pipelines** span rare inherited liver diseases and autosomal dominant retinitis pigmentosa; three liver programs are in in vivo lead optimization
- The company is a **University of Zurich spinout**, with a founding team holding ETH Zurich and UZH doctorates in pharmaceutical sciences, biology, and computational biology
- The seed raise is small relative to platform ambition; the Series A milestone will hinge on demonstrable cycle-time and cost compression versus conventional directed evolution

---

## Frequently Asked Questions

**What is Nerai Bioscience building?**
Nerai Bioscience is developing an AI-integrated high-throughput protein engineering platform that generates novel CRISPR Cas9 variants with customized PAM recognition. By combining automated directed evolution, large-scale cellular screening, and machine learning, the company aims to create editing tools capable of targeting the roughly 80% of pathogenic mutations that existing CRISPR tools cannot reach.

**How much has Nerai raised, and from whom?**
Nerai completed a seed round of CHF 500,000 in May 2025, with Kickfund as the sole disclosed investor. The funds are allocated to preclinical development of the platform and its rare disease pipelines.

**Which diseases is Nerai targeting?**
The company has 8 pipelines: type I citrullinemia (NB-301), phenylketonuria (NB-20X), alpha-1 antitrypsin deficiency (NB-101), homocystinuria (NB-401), classic type I galactosemia (NB-501), primary hyperoxaluria (NB-601), one undisclosed indication (NB-701), and autosomal dominant retinitis pigmentosa (NB-e101). The first three liver programs are in in vivo lead optimization.

**What is PAM restriction and why does it matter for CRISPR therapeutics?**
PAM (protospacer adjacent motif) sequences are short DNA motifs that Cas9 must recognize adjacent to its cut site. Different Cas9 variants recognize different PAMs, limiting which genomic locations any given tool can edit. For rare disease therapy — where pathogenic mutations are scattered across the genome — PAM restriction is a primary reason current tools can only address a minority of disease-causing variants.

**Who founded Nerai Bioscience?**
Nerai was co-founded by Vincent Forster (CEO, PhD pharmaceutical sciences ETH Zurich, ex-Versantis co-founder), Kim Fabiano Marquart (screening platform, PhD biology ETH Zurich), Lukas Schmidheini (engineering platform, PhD biology ETH Zurich), and Sasha Melkonyan (machine learning, MSc computational biology ETH Zurich). It is a spinout of the University of Zurich.