Instrument-free 3D molecular imaging with the VOLumetric UMI-Network EXplorer

Project Information

VOLUMINEX

Grant agreement ID: 101184647

DOI

10.3030/101184647

EC signature date 7 November 2024

Start date 1 March 2025

End date 28 February 2030

Funded under

• The European Innovation Council (EIC)

Total cost

€ 2 999 999,25

EU contribution

€ 2 999 999,25

Coordinated by

KUNGLIGA TEKNISKA HOEGSKOLAN
Sweden

Periodic Reporting for period 1 - VOLUMINEX (Instrument-free 3D molecular imaging with the VOLumetric UMI-Network EXplorer)

Reporting period: 2025-03-01 to 2026-02-28

Summary of the context and overall objectives of the project

Understanding how cells are organized in space is central to modern biology and medicine. Tissues are structured systems where spatial arrangement influences function, disease, and response to treatment. Spatial transcriptomics has emerged as a powerful approach to map gene expression in tissues, but current technologies are limited in scale, cost, and ability to capture three-dimensionality. Most existing methods rely on optical imaging or predefined capture arrays, which constrain resolution, throughput, or accessible volume. There is thus a need for new approaches that measure molecular information across volumes while preserving spatial relationships, and that can scale with the advances in DNA sequencing technologies.

The Voluminex project aims to develop a new class of spatial measurement platform based on molecular network formation. Instead of directly imaging spatial positions, the approach encodes proximity relationships between molecules into networks of DNA interactions that can be read out by sequencing. From this information, spatial organization can be reconstructed computationally. This shifts the problem of spatial measurement from imaging to molecular encoding and data analysis.

The overall objective of the project is to establish an end-to-end framework for generating, reading, and reconstructing such molecular networks in three-dimensional biological samples. This includes the development of chemistries for localized DNA amplification in soft materials, strategies for linking spatially proximal elements into networks, and computational methods for recovering spatial structure from sequencing data. By enabling volumetric spatial analysis without reliance on complex imaging systems, the project aims to expand access to spatial omics technologies and support advances in biomedical research, drug development, and precision medicine. In the longer term, this approach contributes to strengthening European capabilities in sequencing-based technologies and data-driven life sciences.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

During the reporting period, we focused on the core technical foundations for a sequencing-based approach to 3D spatial measurement. The work concentrated on developing experimental and computational building blocks that will support the full end-to-end system.
A primary area of activity has been the development of methods for localized DNA amplification in soft materials, which form the basis for generating spatially confined molecular features. Experimental work has led to the establishment of a robust screening framework to evaluate conditions that promote spatially constrained amplification while suppressing non-specific background signals. This has enabled systematic exploration of key parameters and has produced conditions under which localized amplification can occur. In parallel, computational models have been developed to better understand the physical and chemical factors governing this process and to guide experimental optimization.

In addition to these efforts, initial design work has been carried out toward strategies for forming networks of interacting molecular elements. Conceptual and early experimental exploration has focused on approaches to link neighboring amplified regions in a way that preserves information about spatial proximity. While this part of the system is still under active development, the project has defined a set of candidate architectures and design criteria that will guide further implementation. On the computational side, progress has been made in developing methods to reconstruct spatial organization from network data. Simulation frameworks have been established to generate synthetic datasets that mimic the expected properties of the experimental system. These datasets have been used to train and evaluate machine learning models, including graph-based approaches, which have shown promising performance in recovering spatial relationships under a range of conditions. Current efforts are focused on improving the robustness and generalization of these methods to accommodate variability in network structure.

Activities related to biological integration and application-specific workflows are at an early stage. Initial planning has been carried out to define requirements for future work on biological samples and downstream analysis. Overall, the project has made progress in de-risking key components of the approach by establishing experimental screening capabilities, developing theoretical and simulation frameworks, and demonstrating the feasibility of computational reconstruction from network-based representations.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The approach we are developing in this project aims to change how spatial molecular information can be acquired and interpreted. Rather than relying on direct optical observation or predefined spatial indexing schemes, we explore a strategy in which spatial relationships are encoded into molecular interaction networks and recovered through computational analysis. This introduces a new measurement paradigm that separates the physical process of information capture from the reconstruction of spatial organization. Current spatial transcriptomics technologies are typically limited by trade-offs between resolution, field of view, and cost, particularly when extending to three-dimensional samples. By contrast, a sequencing-based network approach has the potential to scale more naturally with sample size and complexity, while maintaining high spatial fidelity. If successfully implemented, this could allow volumetric mapping of biological systems that are currently difficult to access, such as large tissue sections or organoid models, without reliance on complex imaging infrastructure.

The results so far indicate that key components of this paradigm are feasible. In particular, the establishment of controlled, spatially localized molecular amplification and the demonstration of computational reconstruction from network-like data provide early evidence that spatial information can be encoded and recovered in this manner. While these elements are not yet integrated into a complete system, they indicate a route toward developing this new class of spatial measurement technologies. Beyond immediate applications, our project contributes to a broader conceptual advancement that frames spatial inference as a problem of network reconstruction. This creates opportunities to apply and further develop methods from graph theory and machine learning in the context of molecular data, and to explore hybrid experimental–computational and AI/ML-enabled systems where measurement and inference are tightly coupled.

To ensure further uptake and success, we will need integration of subcomponents into an end-to-end workflow, validation in biological samples, and demonstration of reproducibility and scalability. In addition, translation to practical use will require consideration of manufacturability, protocol standards, and compatibility with existing sequencing infrastructures. This line of work could support tools for studying biological systems, enable accessible spatial analysis across research and clinical settings, and contribute to the development of scalable molecular technologies aligned with emerging needs in data-driven life sciences.