The Signal
Scientific Push vs. Clinical Pull
Two questions decide where a precision-medicine technology really is. How fast is the science still moving — its push? And how far has it been pulled into the clinic — its pull? Plot the two against each other and the field sorts itself into four states, and what each technology needs next becomes obvious.
Bubble size = 2025 publication volume. X = recent publication growth (the adoption model’s median 1/3/5-yr CAGR); the divider is the literature’s ~4%/yr ambient growth. Y = share of 2023–2025 papers in clinical/medical journals (Scimago clinical subject categories), a journal-discipline proxy for clinical translation. Validated against the AI application classifier where available — NGS 42% vs 45%, cfDNA 64% vs 85%, the discovery platforms ~30% vs ~18%: the absolute values differ but the ranking and quadrant placement agree. Dividers are illustrative. This is scholarly translation, not commercial adoption.
Clinical breakout
Still scientifically accelerating and already pulled into the clinic. cfDNA is the clearest case — liquid biopsy, MRD, and early detection are translating while the science keeps moving.
Mature infrastructure
Growth has plateaued, but it is deeply clinical and enormous — the backbone the rest run on. NGS sits here: ~10,000 papers a year, half of them clinical, no longer "hot."
Translation gap
High scientific momentum, low clinical pull — the science is racing ahead of translation. The six discovery platforms (single-cell, spatial transcriptomics, spatial proteomics, long-read, multi-omics, hi-plex) cluster here. The biggest cluster, and the open question: which of them crosses over?
Niche / fading
Stalled and non-clinical. Empty today — no tracked platform is both losing scientific momentum and absent from the clinic. (That can change.)
The shape of the map is the message: the clinically translated technologies have already cooled scientifically, and the scientifically hottest ones have not yet translated. cfDNA is the lone breakout; NGS is the mature backbone; everything else is still science looking for its clinical pull. That gap — between scientific promise and clinical reality — is the whole story of the tools-and-diagnostics layer.
How it’s built
Momentum (x) is each technology’s recent publication growth — the Adoption Index’s median 1/3/5-year CAGR. The divider is the literature’s ambient growth (~4%/yr): above it a field is still outpacing science as a whole.
Clinical pull (y) is the share of each technology’s 2023–2025 papers published in clinical/medical journals (Scimago clinical subject categories) — a journal-discipline proxy for translation that covers all eight technologies consistently. We validated it against the AI application classifier for the four technologies that have one: NGS 42% vs 45%, cfDNA 64% vs 85%, and the discovery platforms ~30% vs ~18%. The absolute values differ — the journal proxy reads the discovery platforms a little high and cfDNA a little low — but the ranking and quadrant placement agree, which is what the map relies on.
Bubble size is 2025 publication volume. Dividers are illustrative, not thresholds. This measures scholarly translation — where the research community publishes — not commercial adoption or reimbursement.