Proposed research scenario
Same cell counts, different neighborhoods
Consider two hypothetical tumor sections with the same proportions of malignant, immune and stromal cells. In one, a rare treatment-persistent state clusters beside fibroblasts; in the other, those cells are dispersed. The research question is whether that local association justifies a different follow-up experiment.
- Same countsStart with matched summaries
Retain cell identity, donor, condition and location.
- Different neighborsTest the local association
Compare observed proximity with a suitable spatial null.
- One hypothesisChoose a perturbation
Test whether the neighborhood supports persistence.
- Supply the measurements that distinguish the sections · Use cell or spot coordinates, RNA and protein measurements, cell-state annotations, donor identifiers and treatment conditions. Preserve the rare population and inspect segmentation, resolution and batch effects before interpreting a neighborhood.
- Produce a checkable spatial result first · Squidpy provides a suitable baseline for neighborhood graphs and enrichment analysis. The planned output is a table of associations by donor and condition, with rare-cell counts and sensitivity to the neighborhood definition. A pooled mean is the simpler comparator.
- Test the proposed BioTwin connection · The proposed connection links a reproducible local association to a stated resistance mechanism and its supporting evidence. A perturbation can then distinguish a fibroblast-dependent effect from a cell-intrinsic explanation. Colocation establishes an association; the perturbation tests the causal hypothesis.
An average can hide the part of the tissue that matters
A program may need to distinguish a resistant niche from a responsive region, or ask whether a proposed cell-state change is blocked only in a particular neighborhood. Location, donor and condition belong in that question.
Squidpy already provides spatial graphs, neighborhood statistics and image-linked analysis. BioTwin’s proposed direction is to connect local biological context with cell-state and intervention hypotheses.
| Question | Familiar approach | BioTwin focus | Practical consequence |
|---|---|---|---|
| Available analysis | Squidpy analyzes spatial molecular data using coordinates, neighborhood graphs and tissue images. | BioTwin currently supports cell summaries; richer spatial integration remains a research direction. | Start the collaboration with a validated spatial-analysis baseline. |
| Research objective | Spatial analysis can identify local relationships and test neighborhood hypotheses. | Investigate donor- and condition-aware neighborhoods that may alter an intervention hypothesis. | Ask whether the same proposed intervention faces different constraints in different locations. |
| Qualification | Data quality and study design determine which biological conclusions the analysis supports. | The proposed integration must preserve rare populations and demonstrate useful downstream decisions. | Evaluate the specific composition before relying on it for program selection. |
The proposed spatial integration and its connection to intervention models are subjects for a research collaboration.
A collaboration with a concrete test
The next useful step is a defined dataset and a decision the proposed integration should improve. Compare the resulting neighborhoods with a suitable established spatial analysis. Check that rare populations survive integration and that any inferred constraint is supported beyond the cell-summary layer.
What a useful connection could look like
Spatial context could identify where a cell-state barrier or modeled escape route is relevant. The connection to these methods remains prospective and should be tested against an independently supported biological observation.
- Cell-state constraints · Ask which barrier is necessary in the relevant setting. Cell-state constraints
- Resistance hypotheses · Investigate whether a local context supports a bypass route. Resistance hypotheses