Why We Invested in Anto Biosciences

At 007 Venture Partners, we invest in companies tackling large, structural inefficiencies hidden in plain sight.

One of the most expensive and under-addressed problems in global healthcare is this:

The human gut microbiome fundamentally alters how drugs work — yet drug development largely ignores it.

This oversight costs the pharmaceutical industry billions annually in failed trials, adverse drug reactions, and geographically inconsistent efficacy.

Anto Biosciences is building the foundation model that makes the microbiome computationally tractable for the first time, transforming drug development from probabilistic guesswork into programmable biology.

The Multi-Billion-Dollar Blind Spot in Drug Development

Global pharmaceutical R&D exceeds $250B annually. Yet more than 90% of drug candidates fail in Phase II/III trials.

A growing body of evidence points to a major driver: the same drug molecule behaves differently across populations because microbiomes differ across geography, diet, age, and lifestyle.

A cancer drug that works in East Asian trials may fail in U.S. trials.
A therapy effective in one demographic may underperform elsewhere.

The industry’s current solutions are slow and expensive:

• Animal models with limited human translatability
• In vitro gut simulators costing millions per study
• Generic computational tools not designed for microbiome complexity

The microbiome is approximately 99% noise. Traditional machine learning struggles to extract signal.

This is the structural bottleneck.

Anto’s Breakthrough: Quality-Aware Sparsification

Anto applies a novel approach called quality-aware, goal-directed sparsification, reducing metagenomic noise by 99% and isolating only predictive signal.

Instead of brute-forcing raw biological data, Anto makes it computationally legible.

The result is a multimodal foundation model for microbial communities that can:

• Predict population-specific drug efficacy
• Identify microbiome-driven toxicity risks
• Simulate drug performance across global subgroups
• Optimise molecules for broader, universal efficacy

What large language models did for text, Anto is doing for microbial ecosystems.

From Molecule Input to Optimised Output

Anto’s workflow is simple but powerful.

Input any drug molecule. The platform returns:

• How different populations will respond
• Where and why it fails by geography, age, or diet
• An optimised molecular variant designed for broader success

This transforms drug development from reactive trial and error into proactive design.

The platform is trained on MetaOmics-10T, one of the largest curated causal datasets for microbial ecosystem modelling, published at ICLR 2026.

The Founders Behind the Platform

Anto is led by:

Arvid Gollwitzer, CEO and Co-founder
Broad Institute researcher who pioneered sparsification methods for microbial data. Published in Nature and at NeurIPS. Built the core algorithms underpinning the platform.

David de Gruijl, Co-founder
Formerly at Janssen Pharmaceuticals, ETH Zurich, and Brigham and Women’s Hospital at Harvard Medical School. Worked directly on drug–microbiome interactions in clinical settings.

This is rare founder-market fit: the scientist who solved the data problem and the clinician who experienced the industry failure firsthand.

What stood out to us beyond their credentials is their grit. Before raising capital and securing pharma engagement, Arvid and David were effectively living in the lab, sleeping there during critical research phases and surviving on bananas while pushing their models to work. They chose to build the hard, scientific solution rather than pursue immediate commercial roles, even though both had clear opportunities to earn significant income within pharma. That level of conviction and long-term orientation is rare, and it shows in both the technical depth of the platform and the seriousness with which pharma partners now engage them.

They are supported by a world-class scientific advisory board spanning Imperial College London, Yale, and UCSF.

A Venture-Scale Business Model

The company already holds $750K in non-binding LOIs and additional pilots in advanced discussions.

The long-term upside is significant. One to three percent of a $5B drug equals $50M–$150M annually per asset.

Even a small royalty portfolio creates venture-scale returns independent of partnership revenue.

Why This Matters Now

Three forces are converging:

• Increasing regulatory focus on population diversity in trials
• Acceleration of cross-border drug licensing
• Pressure on pharma to reduce late-stage failure rates

At the same time, advances in AI foundation models and biological data processing make microbiome simulation finally feasible.

Anto sits directly at this inflection point.

Defensibility: A Data and IP Moat

Anto’s moat is multi-layered:

• Proprietary sparsification algorithms
• MetaOmics-10T training dataset
• Accumulating pharma drug–microbiome interaction data
• Embedded milestone-based LOIs with defined enterprise transitions

Large foundation model players do not have access to curated, causal microbiome datasets at this depth.

Each pharma engagement strengthens the model.

Impact Beyond Returns

Anto enables biologically inclusive drug development:

• Fewer adverse reactions
• More consistent efficacy across populations
• Lower clinical trial attrition
• Reduced time to market for life-saving therapies

This aligns strongly with global health equity and the future of precision medicine.

Why We Invested

Anto represents the type of company we seek at 007 Venture Partners:

• Founders with exceptional scientific credibility and lived domain experience
• A massive structural inefficiency in a $250B+ industry
• Proprietary data and algorithmic IP
• A credible path from pilot to partnership to royalty

Anto is not a point solution for drug trials.

It is the computational infrastructure layer that makes the microbiome programmable for pharmaceutical science.

If successful, it has the potential to redefine how drugs are designed, tested, and optimised globally.

We’re proud to support the team as they build the microbiome foundation model for the next era of drug development.

‍