The market a real-time enzyme-state sensor would serve. Two distinct go-to-markets sit under one technology: (A) enzyme engineering / directed-evolution screening (“find better biocatalysts”), and (B) in-line bioprocess monitoring / process analytical technology (“control the reactor”). They have different buyers, sales cycles, and value-capture. The transduction option that motivates this page is Magnetic Biosensing, because it survives the broth conditions that defeat optical and ISFET sensors.
The two go-to-markets
GTM A: enzyme screening / directed evolution. An instrument that reads per-variant turnover directly (label-free, on a magnetic bead) inside a high-throughput screen, replacing or extending fluorescence/absorbance readouts. Buyer: enzyme-engineering and synbio firms (Codexis, Ginkgo Bioworks, Cradle, EnginZyme, Solugen, Novonesis, Basecamp Research, Adaptyv) and pharma process chemistry. Pain relieved: today screening is the bottleneck, and it is gated to activities you can make fluoresce. A direct turnover read widens assay generality (screen reactions with no fluorogenic substrate) and feeds clean functional labels into AI design loops. This is the stronger “why now.”
GTM B: in-line bioprocess sensing (PAT). A sterilisable in-broth chip reporting enzyme/metabolite state continuously, closing the loop on feed and temperature. Buyer: biomanufacturers and PAT vendors (Hamilton, Mettler-Toledo, Sartorius, Repligen, Endress+Hauser) as an OEM sensor. Pain relieved: in-broth sensing is physically hard. Optical probes foul and cannot see through turbidity; ISFET/electrochemical signals are screened out by the ~0.7 nm Debye length in high-salt broth. A magnetic readout escapes both walls. Today you can measure pH, dissolved O2, biomass and at-line glucose/lactate, but not actual enzyme/metabolite state inline.
Market-size signals
| Market | Size | CAGR | Source (tier) |
|---|---|---|---|
| Industrial enzymes (end-market) | $7.9 to 8.4B (2025) | 6 to 7% | Fortune BI; MarketsandMarkets (2) |
| Process analytical technology (PAT) | $4.0B (2025) to $11B (2035) | ~10.7% | Precedence (2) |
| Biopharma PAT sub-segment | ~$1.0B (2023) to $2.6B (2029) | ~16% | MarketsandMarkets (2) |
| Biocatalysis (pharma/chem slice) | inconsistent $0.74B to $13B (2025), do not quote | 5 to 6% | analyst chaos, flagged unreliable |
| Long-read (single-molecule) sequencing, analogy only | $0.76B (2024) to $3.1B (2029) | ~33% | MarketsandMarkets (2) |
Caveat: none of these is the deal’s true SAM. GTM A’s real market is the enzyme-discovery toolchain (programmes x screening spend), a fraction of the enzyme product market, and must be sized bottom-up. GTM B’s true SAM is the sensors-and-probes slice of PAT, not the whole figure. The “biocatalysis market size” varies more than 10x across analysts ($0.74B vs $13B for 2025): definitional chaos, treat as directional only.
Incumbents / comparables
| Company | Segment | What they do | Status |
|---|---|---|---|
| Codexis | enzyme-eng | CodeEvolver directed evolution; enzymes for Januvia, Paxlovid; ECO Synthesis | Public; FY25 rev ~$64 to 68M; $37.8M Merck deal Oct 2025 |
| Ginkgo Bioworks | platform | Cell-programming foundry, strain/enzyme dev | Public; 2025 rev ~$167 to 187M guide |
| Cradle | enzyme-eng (AI) | Gen-AI enzyme design plus wet lab | $73M Series B (IVP, 2024) |
| Solugen | biomanufacturing | Evolved enzymes plus catalysis for industrial chemicals | Private |
| Advanced Silicon Group | PAT | Si-nanowire photoelectric biosensor for host-cell-protein, <15 min | MIT spinout, grant-funded; closest GTM-B comparable |
| Hamilton / Mettler-Toledo / Sartorius / Repligen | PAT | Incumbent in-line probes (pH, DO, biomass), Raman/NIR | Public/large |
| PacBio / Oxford Nanopore | single-molecule (analogy) | SMRT / nanopore instruments | Public; proves single-molecule can be a product, in sequencing not enzymology |
Why now
- GTM A: in directed evolution the functional-assay step is the bottleneck and is locked to fluorescence. Droplet sorting hits 1 to 2 kHz; even at a million variants per hour, one year covers under 0.1% of a library, and most droplet assays still need a fluorogenic readout. Meanwhile AI enzyme design (Cradle, foundation-model protein work) is demand-pulling for clean functional labels at scale. A label-free turnover read is timely.
- GTM B: PAT measures pH/DO/biomass, not enzymatic/metabolite state inline, and the reason is physical (Debye screening plus fouling), not lack of effort. Magnetic transduction is one of the few classes that escapes both walls.
Where the deal fits (the bifurcation)
- Single-molecule resolution is NOT required for GTM A screening. Ranking variants needs relative per-bead turnover, and ensemble per-bead averaging is enough to order a library. Single-molecule buys mechanistic insight (a research tool), not better ranking. If the pitch leans on single-molecule for screening, that is a flag, not a feature.
- The more credible “sense to manufacture” bridge is GTM A. Enzyme-engineering value is captured by licensing or supplying the resulting biocatalyst (Codexis monetises CodeEvolver enzymes into named pharma processes). The sensor sits upstream of a proven value-capture model: a better screen gives a better enzyme gives a licensable asset. GTM B (closed-loop control) is the larger but slower sale into a conservative, GxP-validated buyer where value accrues to the manufacturer, not the sensor vendor. Higher enterprise value and faster validation on A; larger but slower, more physics-defensible moat on B.
Evidence base
Grounded in six web sources (2026-06-04 pull). Key: industrial enzymes ~$7.9 to 8.4B at 6 to 7% CAGR; PAT ~$4B (2025) to ~$11B (2035) at ~10.7%; droplet screening covers under 0.1% of a library per year and is fluorescence-gated; Debye length ~0.7 nm vs 5 to 10 nm protein explains why charge-based in-broth sensing fails; Advanced Silicon Group is the closest in-broth-sensor comparable. Company name “Six Biosciences” did not surface publicly (only the unrelated Sixfold Bioscience): treat as genuinely stealth and unverified.
Connected ideas
Sits under Synthetic Biology; the transduction layer is Magnetic Biosensing; adjacent to Cell-Free Systems. Entered the KB via the George / stealth magnetic-enzyme deal (see deals/lithic).