This website uses cookies

Read our Privacy policy and Terms of use for more information.

Slush opens in Helsinki on 18 November, and this year we did something unwise: we googled every single company and investor. 2,750 startups, 1,126 investors, 121 biotech and life sciences companies, and €7.1 billion raised. Below are the ones we found cool: a laser that fingerprints dementia, a box that grades antibodies, mini-midbrains by the plate, zebrafish on a conveyor belt, proteins that mine lithium, and alpacas enlisted against sea lice. Then the fund behind a surprising share of the Finnish ones.

Table of Contents

🔦 A laser reads dementia in a drop

RESEARCH & STARTUP

Brainalyze, from Southampton, wants to turn a drop of biofluid into what it calls a biochemical fingerprint. The method is multi-excitation Raman spectroscopy, MX-Raman: hit the sample with lasers at several wavelengths, read how its proteins, lipids and sugars scatter the light, and let a model sort the pattern. The first product line is a dementia triage test that tells a clinic who should go on to imaging or a specialist.

The University of Southampton says the UKRI-funded HOpE project, led by Sumeet Mahajan and neurologist Chris Kipps, has assessed 150 patients so far and is being spun out through Brainalyze Ltd. Initial results put average accuracy above 93% for Alzheimer’s disease, and the grant came from a £32.4 million scheme that drew nearly 900 applications.

Now the methods section. The early work that separated Alzheimer’s from healthy controls used spinal fluid, which nobody volunteers twice. The move to blood is still a registered study: Val-Raman will test the method on stored blood samples from earlier Alzheimer’s and Lewy body cohorts. Southampton itself says validation and miniaturisation come first, with clinical trials after that. Meanwhile, blood tests for phosphorylated tau are already reaching clinics, so a fingerprint has to beat a named biomarker, not just a memory questionnaire.

🧪 Four antibody quality checks, one box

STARTUP

If you make therapeutic antibodies, you spend weeks asking four dull questions: how much did the cells make, what sugars are on it, is it clumping, and is its charge profile right. Each answer normally needs its own instrument and its own expert. Constellate Proteomics, a TU Delft spinout, says one benchtop device can answer all four from a tiny, untreated sample of culture liquid, in minutes.

The company is new enough to still smell of the university. It spun out in February 2026 and in October announced €150,000 in convertible financing from Tech Scout Venture Fund and imec.istart, with Graduate Ventures and Delft Enterprises joining later. It licenses two TU Delft patents, and CEO Carlos de Lannoy did his PhD in single-molecule biophysics, which one of us considers a sign of good character.

Here is the part to read twice. The one accuracy figure in public is for titer: readings within 8% of the true value 19 times out of 20. Titer is the easy question. Glycosylation, aggregation, and charge variants are where mass spectrometry and chromatography earn their keep, and we found no public numbers for those three yet.

🧠 Mini-midbrains in 96-well plates

RESEARCH & STARTUP

Last week we covered three brain organoids sorting Braille. PEARL-Organoids, from Münster, is selling the opposite of a stunt: boring, identical midbrain organoids, delivered in the 96-well plates pharma already screens in. The pitch is batch-to-batch variation under 4%, organoids ready in 35 days instead of 60 to 100, and an automated pipeline that stains and images each one whole.

The description matches a well-documented line of work at the Max Planck Institute for Molecular Biomedicine. In 2020, Jan Bruder’s group published a fully automated workflow in eLife in which 99.7% of wells still held an organoid after 30 days, and Max Planck says its robot can generate, feed, and test up to 20,000 organoids a day, against a few hundred by hand. The work won Germany’s 2021 Animal Welfare Research Prize. The midbrain matters because it holds the dopamine neurons that die first in Parkinson’s disease.

🐟 Zebrafish on a conveyor belt

STARTUP

Most drug discovery starts with a target in a dish and finds out much later what a whole body thinks of it. ZebraMed, a three-person company in Paris, starts with the whole body, as long as the body is a zebrafish larva. Its microfluidic chips use hydrodynamic traps to line larvae up for continuous imaging, behaviour tracking and automated mRNA sequencing, which the company says runs at ten times normal speed. An AI layer then tries to translate fish biology into human predictions.

The logic is sound. A zebrafish larva is a transparent vertebrate with a heart, a liver and a nervous system, small enough to fit in a channel, and most human disease genes have a fish counterpart. Putting them on chips is not new either: a zebrafish-on-a-chip toxicity platform was published in 2014. What is new is treating the fish as a data factory for AI models, and ZebraMed is not alone there. Vancouver’s Zeroshot Bio describes itself as an AI-augmented zebrafish CRO, and there is now a benchtop machine for sorting zebrafish embryos by image.

The evidence is the thin part. We found no paper, preprint, or named customer, only the company’s own account of letters of intent and pharma interest within its first year.

🔋 Proteins that fish lithium out of brine

STARTUP

Not every biotech at Slush is trying to cure something. Magmatic Bio, from Vienna, engineers proteins that bind lithium, grows them by ordinary industrial fermentation, and fixes them into a material that pulls lithium out of messy liquids. The target is direct lithium extraction from sources that today’s methods give up on.

A 2023 pitch sheet explains the prize. Years of evaporation mining have left tailings that hold an estimated one million tonnes of lithium carbonate equivalent, stranded because there is too much magnesium mixed in. The same sheet says the team found a lithium-binding protein and engineered its affinity from millimolar to micromolar, and lists Joseph Cotruvo, whose lab showed that the protein lanmodulin can separate rare earths, as a scientific co-founder. Since then, Rio Tinto has backed Magmatic as one of six critical-minerals start-ups.

The chemistry is the hard part. Lanmodulin works on rare earths because triply charged ions are easy to grip tightly. Lithium is a small ion with a single charge, sitting in brine with vastly more sodium and magnesium, and the protein has to survive that brine, let go on command, and be reused for a long time. No pilot data is public. The idea is in the air, though: the UK’s ARIA is funding a three-year project on protein membranes that separate lithium from sodium.

Enzymes took over laundry detergent on the same argument. The washing machine was a gentler place than a salt flat.

🦙 Alpacas versus sea lice

RESEARCH & STARTUP

There is no alpaca emoji, so the llama is standing in. AquaNab, from Hamburg, immunises alpacas with sea-louse proteins, harvests the small, sturdy antibody fragments that camelids make, and mixes them into salmon feed. The parasite it is after costs salmon farming an estimated $4 billion a year, and Norway’s 2024 season was reported as its worst in more than a decade.

The clever step is the delivery route. Farms currently fight lice with chemical baths and mechanical delousing, both rough on the fish. AquaNab says its nanobodies survive the gut, cross the intestinal barrier and show up in the salmon’s bloodstream, and that they bind louse proteins in the lab. Chief scientist Alejandro Rojas Fernandez built neutralising nanobodies against SARS-CoV-2, CEO Ruth Tamara Montero is a fish immunologist, and Helsinki’s Nordic Foodtech VC has made AquaNab its first investment outside the Nordics.

What is missing is a louse falling off a fish. Binding is not protection, a lesson this newsletter last learned from scorpion toxins. Efficacy trials were only about to start when the company went public in August, and Montero was candid about it: “Working with nanoantibodies and oral delivery is not an easy task.” Then comes the arithmetic of feeding recombinant protein to millions of fish at a price a farmer accepts, and the large animal-health companies chasing the same parasite.

💰 Who wrote the cheques: Lifeline Ventures

INVESTOR

Count the logos on the exhibitor list and one Helsinki fund keeps turning up. Lifeline Ventures has 24 portfolio companies exhibiting, second only to Antler’s 29. Founded in 2009 by Timo Ahopelto and Petteri Koponen, it is best known for early cheques into Supercell, Wolt and Oura. Most of its Slush contingent makes furnaces, concrete or software, but four of them touch the lab bench, and they sit at very different distances from a customer.

CurifyLabs is the furthest along. Its machines let pharmacies print personalised doses of medicine on site, and in July it closed a $14 million Series A co-led by Sandwater and HealthCap, with pharmacies in 21 US states already using the system. Measurlabs was started by two PhD students tired of hunting for test labs, and now routes orders to more than 900 laboratories through one platform. DEEP Measures runs a shared platform for building digital measures, such as wearable readouts, into clinical trials. Elea & Lili, a VTT spinout, raised €2.5 million in March, led by Lifeline, to replace the fossil-based superabsorbent in diapers with cellulose.

And then the one that is not at Slush. Disclosure: Prateek is founder of Proteins.1, so read this as a humble brag. Lifeline co-led the company’s €4.7 million pre-seed in April with Cloudberry. The technology reads a single captured protein again and again with magnetic cycling instead of an enzyme. The claim is sensitivity up to 1,000 times beyond today’s best platforms. “Up to” is carrying weight in that sentence, and we would say the same about anyone else’s press release.

It seems like Lifeline writes the first cheque when the evidence is one lab and a patent, then waits.

We usually spend this space checking headlines against methods sections. This week there were mostly no methods sections, just announcements, which are a genre with their own conventions.

If you are in town, the life science founders pitch from the Helsinki SkyWheel on 17 November, and Y Science, the signature life science side event, expects 1,500 people. Which booth would you visit first? Hit reply, we read everything. And if you know someone flying to Helsinki in November on purpose, forward this along.

Keep questioning everything (especially anything written on an application form),

P.S. The sun sets in Helsinki at about half past three on the day Slush opens. More than 12,000 people fly in anyway. We are told this is the point.