We are a Tel Aviv-based company, so this perspective is not disinterested. But ecosystem claims deserve scrutiny, not marketing copy — so let us try to make the case for what Israel actually offers computational biology, rather than what press releases say it offers.
The honest answer is: a specific, dense concentration of people who can write code, read genomes, and understand biology simultaneously. That combination is rarer than it sounds, and the Israeli research-to-military-to-startup pipeline has been producing it at an unusual rate relative to population size.
The Academic Foundation Is Unusually Strong
Three institutions anchor the pipeline. The Weizmann Institute of Science in Rehovot has historically been one of the strongest bioinformatics programs globally — the department has produced multiple foundational papers in sequence alignment, protein structure prediction, and comparative genomics that have shaped how the field thinks about problems. Weizmann graduates are not trained as biologists who learned to code; they are trained as computational scientists with deep biological domain knowledge, which is a different cognitive profile and produces a different kind of researcher.
Hebrew University's Faculty of Medicine and School of Computer Science produce researchers who sit comfortably across that divide, and the Jerusalem institutions have had a particularly strong output in structural biology, systems biology, and immunoinformatics. The joint degree programs that emerged in the 2010s created graduates who had absorbed both sides of the problem before finishing their PhDs.
Technion, the Israel Institute of Technology in Haifa, anchors the engineering side — strong in machine learning, signal processing, and the mathematical foundations that underpin modern sequence analysis. When deep learning entered biology in force around 2016-2018, Technion researchers were already fluent in the methods; the translation lag between ML advances and biological application was shorter in Israel than in many Western academic environments.
None of this is unique to Israel — MIT, Stanford, Cambridge, ETH Zurich all have strong programs. The difference is concentration. In a country of roughly ten million people, that density of bioinformatics expertise in a single metropolitan corridor from Tel Aviv to Jerusalem to Haifa is remarkable. Researchers cross-pollinate frequently, and the academic community is small enough that people know each other across institutional boundaries.
The Spinout Record: Compugen, Immunai, MeMed
Compugen is the origin story the ecosystem points to. Founded in the late 1990s by researchers applying algorithmic biology to gene discovery, Compugen demonstrated in the pre-genomics era that computational approaches could generate biologically meaningful hypotheses. The methodology was ahead of what experimental throughput could validate at the time, but the framing — use computation to identify candidate targets, then hand to wet lab — became the template for what Israeli biotech would do well.
Immunai, a 2018 venture from alumni with Weizmann and MIT connections, applies single-cell genomics and ML to map the immune system at resolution that was not previously accessible. Their focus on multi-omic immune profiling represents a natural evolution of the computational biology tradition: not just sequence analysis but integration of transcriptomic, proteomic, and epigenomic layers. The company has moved from platform to pipeline, which is the harder and more interesting transition.
MeMed built something different — a host-response diagnostic that distinguishes bacterial from viral infection using a protein signature rather than pathogen detection. The insight is computational: you can read immune response profiles rather than hunting for the pathogen directly. That is a very Israeli way of framing the problem, prioritizing the signal extraction problem over the traditional microbiological approach.
These three companies span three decades and three different problem areas, but share a common intellectual lineage: compute-first, biology-second in the discovery phase, with wet lab validation as confirmation rather than discovery.
Unit 8200 and the Data Science Pipeline
This part of the story is less discussed in polite company, but it is real and worth naming directly. Israel's military intelligence units — Unit 8200 in particular — have functioned for decades as an intense training ground for signal processing, large-scale data analysis, anomaly detection, and adversarial system design. Conscription means that many of the most technically capable young people in the country spend two to three years working on signal-dense, high-stakes computational problems before entering university or industry.
The translation from signals intelligence to bioinformatics is not obvious, but it turns out to be real. The mathematical foundations overlap — sequence analysis, pattern recognition, working with noisy high-dimensional data, building systems that surface relevant signals from large amounts of irrelevant noise. Researchers who spent time in technical military units and then entered bioinformatics programs often describe a familiarity with the problem structure even if the domain was new.
We are not saying military intelligence training produces better biologists. It does not, directly. We are saying it produces a particular density of people who are comfortable with large-scale data processing, rigorous about signal-versus-noise, and accustomed to working under constraint — and some fraction of that population ends up in computational life sciences. The pipeline is a feature of Israeli society that has downstream effects on the startup ecosystem whether or not anyone explicitly designed it that way.
What This Means for Preclinical AI Drug Design
For a company building an AI-first target identification platform, being in Tel Aviv means something practical: when we look for computational biologists with deep genomics experience, there is a local pool to draw from. That may sound trivial but it is not. Recruiting for a role that requires simultaneous fluency in genome-scale alignment methods, phylogenetic analysis, and machine learning feature engineering is genuinely difficult in most locations. The Tel Aviv corridor reduces that friction.
It also means academic collaboration is accessible. The distance between our offices and Weizmann is about 40 minutes on the road. Conversations with researchers who work on aging genomics, comparative biology, or mRNA biology at Israeli universities can happen in person, and several of our early methodological choices were informed by those conversations rather than by literature alone.
The university-to-startup knowledge transfer in Israel is also notably faster than in some European academic cultures. IP commercialization through technology transfer offices is well-established, researchers are culturally comfortable moving between academia and industry, and there is no strong stigma attached to applying research to commercial ends. That culture shortens the time between a computational insight and a testable therapeutic hypothesis.
The Gaps: What the Ecosystem Does Not Yet Have
Honest accounting requires noting what is missing. Israeli biotech has historically been stronger in the platform and discovery phase than in late-stage clinical development. The clinical trial infrastructure, the patient advocacy networks, and the regulatory relationship-building required to move a drug through Phase II and III have not been built to the same depth as in the US or Western Europe. Companies that do discovery work in Israel frequently shift their clinical operations and regulatory strategy to the US or UK as they progress, which reflects a real structural asymmetry.
Manufacturing capacity is another gap. For mRNA therapeutics in particular, lipid nanoparticle formulation and GMP manufacturing are not areas where Israel has strong local infrastructure relative to, say, Germany or Massachusetts. That means a Tel Aviv mRNA company almost certainly needs CRO and CDMO relationships outside the country for any serious preclinical or clinical manufacturing work.
The ecosystem also has a concentration risk: many of the strongest computational biology graduates are competing for talent at a small number of well-funded companies, which drives compensation pressure that can be challenging for early-stage organizations. The talent pool is real, but it is also contested.
Why We Are Here
caVos was founded in Tel Aviv in 2023 because the founding team was already here, and because the local concentration of computational genomics expertise made building the target identification side of the platform feasible at a stage when we had limited resources. We have been direct about the gaps: our mRNA synthesis and formulation work requires external partners, and any future clinical development will require a regulatory presence in the US or EU.
But the computational biology core of what we do — multi-species genome alignment, conservation scoring, phylogenetic weighting, target prioritization — that work is genuinely well-supported by the local ecosystem. It is not that Tel Aviv is the only place where this could be built. It is that it was a rational place to start, and the density of domain expertise has mattered at every stage of building the platform so far.
The Israeli computational biology scene is serious. It is also specific: strong on discovery-phase computation, weaker on clinical execution, excellent at the kinds of algorithmic biology problems that make up the front half of the drug discovery process. For a company working on that front half, that fit is not accidental.
