reef.bio

The GMP automation platform for personalized therapeutics manufacturing at the point of care.

Reef begins with directed phage evolution against patient isolates of antibiotic-resistant pathogens.

See it in actionSee the Story
Why it matters ↓
AnthropicNVIDIAGooglePlasmidsaurusOpenAI

Drug-resistant infections are winning.

Resistance is not a future problem waiting on a new drug. It is already the background condition of hospital medicine, and the supply of replacements has been thinning for two decades.

  • 35,000 people a year die of drug-resistant infections in the United States, on CDC figures.
  • 1.27 million die worldwide each year, on the Lancet estimate.
  • The antibiotic pipeline is nearly dry, with few genuinely new classes coming.
  • Resistant bacteria do not fade once the drug is withdrawn. In a fluctuation test on E. coli, none of twelve resistant mutants showed a detectable growth cost against the susceptible parent, so nothing pushes them back out.
  • Every new drug restarts the same clock, because the bacteria that survive it are the ones that go on to reproduce.

Sources: CDC Antibiotic Resistance Threats Report; Murray et al., The Lancet 2022;399:629; Chaudhry et al., PLoS Biology 2018;16(8):e2005971.

SINCE YOU OPENED THIS PAGE
0

people worldwide have died of a drug-resistant infection. That is about one every 25 seconds.

NEXT IN 25s · BASED ON 1.27M DEATHS/YR (LANCET)

The data is clear: a fixed drug can’t keep up with an infection that keeps changing.

10²10³10⁴10⁵10⁶10⁷10⁸10⁹036912151821treatment startsBacterial load (CFU/mL)Time since treatment (days)
AntibioticFixed phageMyx, evolving phage
Figure 1. Bacterial load over 21 days under three treatment strategies. Schematic curves drawn to the dynamics reported in the cited sources; axis values show the orders of magnitude involved rather than measurements from a single study.
Antibiotic

The drug works until the bacteria change.

An antibiotic knocks the infection back, then the survivors regrow and the same drug stops working. The survivors are the ones the drug selected for, which is why the second course rarely works as well as the first.

2.3 to 2.7 × 10⁻⁵
That is how often resistant mutants arose per generation in a fluctuation test on E. coli. None of the twelve mutants tested showed a detectable growth cost against the susceptible parent, so nothing removes them once the drug is gone.
Chaudhry et al., PLoS Biology 2018;16(8):e2005971
1 of 3

A patient-matched phage, evolved and produced on site.

Read the storyTake the machine apart

A clinician sends the isolate that is not responding. Myx grows it, puts a phage population under selection against that strain, and keeps the rounds running while the bacteria evolve to escape, so the phage is still being pushed as the target moves.

What comes out is sequenced and release-tested before it goes anywhere near a patient, and it is made in the building where the patient is rather than shipped from a specialist centre weeks later.

Bring us the infection nothing else can treat.

Clinicians, hospitals, investors, and advisors each have a path. Tell us where you fit and we will take it from there.

CliniciansHospitalsInvestorsAdvisors