Skip to content

James Sharpe — "C3PO: Cell 3D Positioning by Optical encoding"

Affiliation: EMBL Barcelona (head of unit; Multicellular Systems Biology) Day / session: Day 2 (Thu Sep 18, 2025) — morning keynote Talk title: "C3PO: Cell 3D Positioning by Optical encoding" Slides (Zenodo DOI): to verify after publish Video: Bioconductor YouTube — plenaries playlist (keynotes recorded) Abstract / schedule: eurobioc2025 schedule

Thesis

C3PO appears to be a new optical-encoding approach to assign 3D spatial positions to individual cells — a non-trivial problem for dissociated single-cell experiments where the tissue context is lost at sequencing time, and a complementary problem for in-situ methods where segmentation in 3D is hard. Sharpe's lab has historically argued that acquisition geometry is itself a methodological frontier (Optical Projection Tomography, mouse limb development atlases), so the talk likely frames C3PO as another "encode the geometry into the measurement" play rather than reconstructing geometry post-hoc.

Speaker context

James Sharpe heads the Multicellular Systems Biology unit at EMBL Barcelona and is one of the inventors of Optical Projection Tomography (OPT), a foundational 3D imaging method for developmental biology. His group's signature work has been on mouse limb morphogenesis — combining live imaging, computational modeling, and quantitative cell tracking. Inviting him to a Bioconductor / methods conference signals that the program wants the upstream end of the spatial-omics pipeline represented: the data that downstream R/Python frameworks ingest is shaped by what the microscope and the labeling chemistry can encode. C3PO is, on the title alone, a continuation of that lineage.

Connections to the corpus

  • SpatialData — at the data-substrate level: any 3D-positioning method has to produce coordinates that downstream frameworks (SpatialData, SpatialExperiment) can hold and visualize
  • AACR axis: single-cell-spatial-omics — but this is the acquisition end of that axis, not the analysis end; most AACR spatial dossiers are about analysis tools that assume coordinates are already in hand
  • Closest analytical kin in the corpus: cross-modality registration / segmentation tooling rather than gene-level methods

Open questions

  • Is C3PO an optical / hardware method, a barcoding-based method, or a hybrid (e.g., spatial barcoding patterns read out by imaging)? The acronym "Optical encoding" hints at hardware-side, but the field has converged on hybrids.
  • What is the readout — can it be combined with single-cell sequencing on the same cells (a la Slide-seq / Stereo-seq lineage), or is it imaging-only?
  • Is there an open-source software component (a Bioconductor or scverse package) that ships with the method, or is this a pure-method talk in front of a software audience?