Amber
Bio.

On any given Tuesday in 2026, a bench in Cambridge is doing something most of biotech has been told is impractical. The team at Amber Bio is taking a strand of RNA and rewriting thousands of bases in a single pass - not flipping an A to a G, not pasting in a short patch, but rewriting whole functional regions. The instrument is small. The reagent looks like every other reagent. The implication does not.

Gene editing has spent the last decade getting more precise and more narrow. Base editors flip a letter. Prime editors swap a short string. Each is a marvel; each is also a single shot at a single mutation. That works for the diseases where one mutation accounts for most patients. It does not work when a disease comes in a thousand flavors - when a single gene can break in a thousand ways and each broken version belongs to a few unlucky families.

Amber Bio thinks the answer is to stop editing letters and start editing pages.

That is the bet. It is a serious one. It is also, given the founders, not a wild one.

Most rare diseases are not actually that rare. Inherited retinal degenerations, for instance, affect hundreds of thousands of people - but no two families necessarily carry the same mutation. Build a drug for the most common variant and you serve a slice. Build a drug for each variant and the economics collapse before you get to the rest.

Amber Bio's platform is aimed squarely at that gap. If a single therapeutic can address thousands of distinct mutations in the same gene, the unit economics of rare disease change. So does the answer your geneticist gives you when you ask, "is there anything for my variant?"