Deep Piction, an EIC Accelerator Winner, Tracks Drugs Cell by Cell

Late-stage drug failures are mostly put down to unforeseen toxicity, off-target effects and lack of efficacy, and MRI and PET, in the words of one EU project, "lack the resolution to provide a comprehensive, single cell view" of where a therapy goes. That project is Deep Piction's own. The company, founded by the Helmholtz Munich researcher Ali Ertürk and based in Neuherberg near Munich, makes a whole mouse transparent, images it in 3D with a light-sheet microscope and uses AI to quantify where a labelled therapy went, cell by cell. Its CORDIS project summary calls the method "zero-slicing, entire-organism coverage, single-cell detail".

The EIC Accelerator selected the project at its 1 October 2025 cut-off on grant-only terms, and CORDIS records an EU contribution of €2,454,812.32, signed on 29 April 2026, for May 2026 to April 2028. With six months of the grant gone, what can be checked is the science behind it, which is published, and the commercial record, which is close to empty: the company names no customer and its news page carries only placeholder entries.

Seeing a whole animal at cell level

The method comes from Ertürk's academic laboratory. He is research director at the Institute for Intelligent Biotechnologies at Helmholtz Munich and a faculty member at LMU Munich, according to LMU's Munich Center for NeuroSciences and his Wikipedia entry, and his group published the DISCO family of tissue-clearing methods, including uDISCO in Nature Methods in 2016 and the panoptic imaging work in Nature Neuroscience.

The company's website describes the pipeline in five steps: a labelled drug is given, the whole body is cleared and imaged by light-sheet microscopy, AI builds an anatomical "digital twin" onto which the signal is mapped, a robotic module samples regions of interest for proteomics and transcriptomics, and the results are turned into pharmacological readouts. Against the established tools, the pitch is resolution: the CORDIS summary argues that the gap left by MRI and PET matters most for mRNA drugs, gene therapies and nanocarriers that must reach specific cells.

The strongest public evidence is a paper. "Nanocarrier imaging at single-cell resolution across entire mouse bodies with deep learning", published online on 14 January 2025 and in the December 2025 issue of Nature Biotechnology, lists Deep Piction among the authors' affiliations, according to the Technical University of Munich's research portal. Its pipeline, SCP-Nano, detected lipid nanoparticles at doses as low as 0.0005 mg/kg, which the authors call far below the limits of conventional whole-body imaging, and it worked on liposomes, polyplexes, DNA origami and adeno-associated viruses. The author list on Weill Cornell's VIVO record includes several people the company now names as staff, among them Ceren Kimna, Harsharan Singh Bhatia and Denise Jeridi.

A grant-only award in a blended round

The cut-off drew 923 full applications asking for almost €7 billion, the EIC reported on 17 October 2025. On 17 February 2026 the EIC announced 61 selected companies from 121 interviewed, with 85% on blended finance. The EIC's list of selected companies shows DEEP Piction GmbH with "Grant Only" as its financing type, one of eight such companies, so the award carries no EIC Fund equity.

The grant itself is narrow in what it promises. With the EIC money, the CORDIS summary says, the company will "finalise the standardisation of mRNA labeling in entire adult mouse bodies, scale our commercial efforts, and remain the only solution offering such transformative 3D subcellular analysis". That last part is the company's claim about its market; web searches for this article found no independent comparison of competing whole-body methods to test it. CORDIS gives no total project cost and, this early, no periodic report.

1. Standardised mRNA labelling in whole adult miceToo earlyProject runs to April 2028; no result published by the company or in CORDIS.
2. Scale commercial effortsNo public evidenceThe website names no customer or partner company; its news page holds three placeholder entries dated March 2026.
3. Cell-level whole-body imaging of nanocarriersPublished before the grantSCP-Nano, Nature Biotechnology, January 2025.
4. Patent portfolio from the academic workNot listedThe CORDIS summary mentions patents "transferred from our academic roots" but names none.

From imaging service to therapy developer

The project CORDIS describes is a service for others: the company "empower[s] biopharmaceutical companies to optimise therapeutic molecules". The website, checked on 4 November 2026, frames the business differently. It says Deep Piction "combines a proprietary lipid nanoparticle technology for extra-hepatic delivery with AI-quantified whole-mouse imaging" and applies both "to the development of in vivo cellular therapies", and it describes a programme for repeated dosing of targeted nanoparticle carriers. No drug candidate, target or disease programme is named on the site.

The two framings are not exclusive: the imaging is the tool in both, and the website still lists biodistribution and target engagement work. But a company that develops its own therapies needs far more capital and time than one that sells imaging studies, and Ertürk told Forbes Türkiye in September 2025 that he wants to move the company to the United States and build it into the world's largest biotech company.

Money reported, not announced

The only figures for the company's finances come from a single interview. Forbes Türkiye reported on 4 September 2025 that Koç Holding had invested €3 million at a €20 million valuation the month before, that a German federal research ministry programme had put in €1.5 million, and that the company had earned €1.3 million since its founding. None of these figures has been announced by the company or found in a second source for this article. The company's team page does list an advisory board member "representing Koç Holding A.S.", which supports the link but not the amount. Koç Holding's August 2026 disclosure of its holdings does not name Deep Piction, but it covers only direct stakes.

Europe's case: fewer animals, better-targeted drugs

The project sits in two EU agendas at once. CORDIS tags all of the grant to the digital agenda and to artificial intelligence. And the company's argument, that therapies which cannot be tracked in detail lead to late failures and "greater reliance on animal testing", lines up with the Commission's roadmap towards phasing out animal testing for chemical safety assessments, published on 1 June 2026 and covering pharmaceuticals among 15 domains, as reported by RAPS and Cosmetics Business. RAPS reports more than 15 million animals used for EU regulatory testing between 2015 and 2023.

The fit has a limit: the method, as published, images whole mice, so it uses animals. The company's stated aim is to reduce animal use by moving towards in silico models, per Wikipedia and the Forbes Türkiye interview; no figure for any reduction has been published.

The reporting is based on publicly available sources, listed below.