Gajus
Worthington

Walk into Superfluid Dx in South San Francisco and the pitch is almost absurd in its ambition: take a tube of blood, fish out the messenger RNA that brain cells quietly shed into the bloodstream, and read it like a transcript of what the brain is actually doing. Gajus Worthington runs the company that thinks this is the future of diagnosing Alzheimer's.

The bet is specific. Existing Alzheimer's tests chase amyloid beta and tau - the plaque-and-tangle proteins everyone has heard of. Worthington's team argues that approach misses up to a third of developing cases and misdiagnoses another third. So they went after a different signal entirely: cell-free messenger RNA, a direct readout of pathology rather than a proxy for it. It is the kind of contrarian, manufacturing-hard idea he has spent a career chasing.

He is not a biologist. He is a physicist and an electrical engineer who learned to treat biology like a fabrication problem - and that lens is exactly what makes his story unusual.

In 1999 he co-founded a company called Mycometrix with Steve Quake, then a young scientist whose Caltech lab had figured out something strange: you could build the equivalent of an integrated circuit, except the channels routed liquids instead of electrons. Microfluidic large-scale integration. A chip for biology. By 2001 the company had a better name - Fluidigm - and a mission to manufacture devices nobody else knew how to make.

The instruments sold for roughly $200,000 each, and the company became, by 2009, described as the world's leading maker of microfluidic devices. None of it was smooth. An IPO attempt in 2008 collided with the financial crisis and failed. The company piled up nine figures of accumulated deficit. Worthington stayed. In February 2011 Fluidigm tried again and raised about $75 million.

He had come to all of this from semiconductors. From 1994 to 1999 he worked at Actel Corporation, a public chip company, rising to Director of Strategic Marketing and Product Planning before leaving to bet on fluids. The semiconductor habit of mind - yield, process, scale - never left him.

There is a pattern in the work, and it is not subtle. Worthington is drawn to the technology that is hard to manufacture - the thing other people sketch on a whiteboard and abandon because the yield is impossible. Fluid circuits. RNA fished out of blood. He treats biology with the discipline of a chip fab, and he stays through the ugly part of the curve. The 2008 IPO failure did not move him off the project; the deficit did not move him. That is rarer than it sounds.

The other tell is who he keeps working with. The collaboration with Steve Quake has now spanned the better part of three decades and two companies. People who are easy to work with attract that; people who are merely brilliant do not.

He is an elected member of the American Institute for Medical and Biological Engineering - a small recognition that hides a large point. The engineer's medal, awarded for the biology.