Paragon Genomics sells the pickaxes of precision sequencing.

Here is a fact about genomics that does not make it into press releases: the sequencing machine is often the easy part. You can buy a sequencer. What you cannot buy off a shelf - or at least, what you could not easily buy in 2015 - is a reliable way to prepare a difficult sample so the sequencer reads only the regions you care about, cleanly, cheaply, and without drowning your data in noise. That preparation step, called targeted enrichment and library prep, is unglamorous chemistry. It is also where a lot of experiments quietly die.

Paragon Genomics is a company built on that boring, load-bearing step. Founded in 2015 in the San Francisco Bay Area by Tao Chen and Zhitong Liu, and now headquartered in Fremont, California, it makes reagent kits - not machines - for targeted next-generation sequencing. Its flagship technology is called CleanPlex, and the pitch is straightforward enough that you can explain it at a dinner party: take a tiny or degraded DNA or RNA sample, and turn it into a sequencing-ready library in about five hours, with much less of the background junk that makes older methods frustrating.

The clever bit - and the thing that makes CleanPlex a product rather than a protocol - is a patented "background-cleaning" chemistry. Amplicon-based methods, which use multiplex PCR to copy the regions of interest, are fast and work with very little starting material. But they generate non-specific PCR products - noise. The competing approach, hybrid capture, is cleaner but slower and fussier. Paragon's scientists decided they did not want to pick. CleanPlex pairs a proprietary primer-design algorithm and a uniform amplification chemistry with a step that scrubs away the non-specific products, aiming for the speed of amplicon with cleanliness closer to capture.

If that sounds like an incremental improvement, consider who buys it. Paragon says more than 300 organizations globally use its panels, and the customer list reads like a directory of serious science: the Broad Institute, MD Anderson Cancer Center, Mayo Clinic, St. Jude Children's Research Hospital, Boston Children's Hospital, Novartis, plus government and academic labs at the University of California, the USDA, the CDC and the Chinese Academy of Sciences. When your reagent shows up in the methods sections of that many good papers, you are no longer selling a chemistry kit. You are selling a dependency.

The business model is the classic "sell the pickaxes" play, and it is a good one. Paragon does not sequence anyone's genome and does not sell instruments. It sells consumable reagent kits and custom design services, ordered B2B - including through its own website - to research institutions, clinical labs, diagnostics companies and agricultural genomics groups. Consumables get reordered. Dependencies compound.

The capital story is refreshingly small. In September 2017 the company raised an oversubscribed $8 million Series A led by Cowin Venture Capital and Fosun Industrial, with participation from Cloudstone Venture, HEDA Ventures and Wisemont Capital. That is, as far as public records show, the extent of the outside funding - no billion-dollar valuation, no hype cycle. With a team of roughly 16 people and estimated revenue around $4.5 million, Paragon looks like what a focused life-science tools company is supposed to look like: modest headcount, real customers, a defensible product.

Paragon is not alone in the target-enrichment aisle. It competes with much larger names: Illumina's AmpliSeq, Twist Bioscience, IDT, Agilent's SureSelect, QIAGEN, and Roche's KAPA products. The interesting question for a company this size is not whether it can out-spend those firms - it cannot - but whether a focused, well-engineered chemistry that handles the hard cases can keep winning labs one validated experiment at a time. So far, 300-plus of them suggest it can.

There is a tidy lesson buried in all of this, and it is the kind that outlasts any single product. The durable businesses in a gold rush are usually not the prospectors. They are the people selling the tools - the reagents, the kits, the boring middle of the workflow that everyone needs and nobody wants to build themselves. Paragon Genomics found one of those middles and made it a company.