The problem. The three imaging spatial platforms — 10x Xenium, Vizgen MERSCOPE, NanoString CosMx — all promise single-cell-resolution, in-situ transcript detection, and vendors all show their best case. The question a facility actually faces is empirical: on the same tissue, how do they compare on sensitivity, specificity, and cell calling?

The idea. A systematic head-to-head on matched FFPE samples. It compares the platforms on the metrics that decide real experiments: transcript detection sensitivity (counts per cell), specificity / background (false-positive signal), the number of cells recovered and segmentation quality, and concordance with reference measurements. Crucially it holds tissue and target genes as fixed as possible so the differences it reports are platform differences, not sample differences.

Why it matters. This is exactly the “which assay should we buy/run for this study?” decision a spatial core facility makes constantly, and it’s the kind of standards-focused, orthogonal-comparison work that a data-and-reproducibility person can own. Reading it is how I’d form a defensible opinion instead of repeating vendor claims — and being able to say “it depends on your tissue and panel, and here’s the benchmark” is the credible answer.

Verdict. A benchmark’s value is its methodology, and the honest caveats are the usual ones: results depend on the specific tissue, panel, and software versions, and the field moves fast enough that today’s ranking can shift with a chemistry update. So the durable takeaway is the framework for comparison, not a permanent winner. My plan: turn it into a one-page decision guide, updated as chemistries change.