Special Edition Virtual Pub: Spatial Proteomics

Norman opened the session by tackling every facility's favourite headache: what to do with enormous high-plex datasets. He introduced Katana Edge, a cloud-native platform that lets teams analyze ultra-high-plex whole-slide images and TMAs directly in the browser – no in-house compute cluster required. AI workflows handle cell and tissue segmentation, phenotyping and clustering across mIF, spatial proteomics, spatial transcriptomics, FISH and IHC. Bonus points for the LLM-powered option to simply ask your data questions.

Knowing where a protein sits is one thing; knowing whether it is actually doing anything is another. Flavia showed how QF-Pro® adds that functional layer, using amplified FRET-FLIM to quantify protein–protein interactions and post-translational modification states in fixed cells and FFPE tissue. The result is a single QF-Pro® Score that reflects how engaged two partners really are. Because it runs independently of other omics workflows, it can be layered onto existing spatial pipelines – with an eye on future companion diagnostics.

David set a course for the final frontier – quite literally, opening with a Star Trek-inspired mission statement "to boldly go where no antibody has gone before." Back on Earth, he asked an honest question: if spatial proteomics relies on fairly simple tools and common skills, why do so many labs struggle to make it work? He shared how Newcastle's facility-led approach helps users navigate the time, cost and complexity, delivering results across a wide range of tissues, targets and diseases.

Gijs took us from pixel to proteome by coupling AI-guided pathology with Deep Visual Proteomics. Foundation models first located precursor lesions in pancreatic cancer tissue; laser microdissection and mass spectrometry then quantified more than 9,000 proteins from around 100 cells. The headline finding: proteome-level reprogramming happens beforethe tissue looks any different under the microscope. Scaled up as tileDVP, the approach generated over 2,500 proteomic measurements across 55 FFPE blocks, building an atlas from earliest precursor to metastatic disease.

Eric closed the session with the big picture. He presented Visual Omics, the next-generation integration of NL-BioImaging (home of the Advanced NL microscopy Euro-BioImaging Node) and the Netherlands X-omics Initiative. The goal: move beyond static snapshots to "record" the molecular history of living cells by linking microscopy and omics facilities within and between institutes. He outlined the challenges to tackle at national level, from smart microscopy that automatically picks out cells for omics analysis to truly cross-modal data integration.