Skip to main content
Premium Trial:

Request an Annual Quote

Through SomaLogic, Gold Hopes to Make a Difference in Medicine



NAME: Larry Gold

AGE: 60


PRIOR EXPERIENCE: Professor of Molecular, Cellular, and Developmental Biology, University of Colorado, Boulder; co-founded Synergen; and founded NeXagen, which later became NeXstar Pharmaceuticals

In the second installment of a two-part interview, Larry Gold talks about the evolution and future of SomaLogic.

 Since you started SomaLogic what has changed?

Well lots of things have changed. I’ve gotten two years older! First of all the idea is photoaptamers, not aptamers, and the magic of that idea is that you get the specificity of a sandwich ELISA with antibodies out of a single capture agent. The core of this company is that a single agent — an aptamer aimed at protein x — can actually bind that protein very specifically in a mixture. When you crosslink that protein in the second step to the aptamer, the specificity gets even higher because things gratuitously adsorbed to the aptamer will not crosslink. The data so far support that quite dramatically. That advantage we understood, actually, a long time ago.

We also understood a long time ago that once you have a protein covalently bound to an aptamer, which is covalently bound to an array, you get to wash that array very fiercely. We’ve been using NaOH, SDS, high salt, low salt, organic solvents, and today in the middle of a discussion someone said, ‘Gee, we could wash these things with acid also.’ And everybody said, ‘Oh yeah, of course, covalent is covalent.’ So now the washing will be under even more harsh conditions. When you’re done washing if you’ve really only got one protein sitting on a feature that’s got a lot of that photoaptamer on it, you ought to be able to quantify how much protein is there by staining the lycines, for example, and measuring those things against the background of aptamers that don’t stain with lycine-specific reagents. We started the company quite some time ago and that has stayed the same. We continued [working] on the robotics to make sure we could make lots of photoaptamers. We’ve built this wonderful robotics capacity to select photoaptamers and we can make very large chips, that is, we can make photoaptamers against proteins at a pretty high level. All of that was expected.

The biggest change that has happened since we concocted this business idea — now six years ago I guess — is that the world has made measurements that support our business intention. There are now enough beautiful papers that show that protein levels go up and down with disease to really think that you can change diagnostics. You’re allowed to hunt broadly and stupidly for surrogate markers, that’s what has changed. I love that. People like us trying to do this in a high-density format could really add some value for pharma and biotech as well as for diagnostics.


What’s the scale at which you can make photoaptamers?

We have two robots at the company, and we think that given our high-throughput methodology to select photoaptamers we can do in the course of a year 10,000 different proteins. One of the limitations in our business model is that we have to get our hands on proteins. We have a few deals, [such as] one that was announced with Celera a half year ago. When Celera finds proteins that are interesting, they give us the protein and we make a photoaptamer. They get to use the photoaptamer for stuff that’s within our license from Gilead, and we get to put the photoaptamers on a chip. We think when people understand what we do more, then we’ll get our hands on more proteins than we now have. That’s our business model.


Ultimately is the goal to sell access to the platform?

No, the goal is to spend all our money! We’re actually a not-for-profit-charity! No, the idea is that sometime early next year or so we will be able to have something useful in the research space for pharma, biotech, and academic people, and to follow that as quickly as we possibly can with small panels and large panels for real clinical diagnostics — the kind of thing done by reference labs [and] the sort of thing that FDA would eventually approve. The idea is to make bigger and bigger and more and more informative chips, and to figure out the right way to approach FDA so that the diagnostic information is of real value to the physician and the patient. We have a very strong informatics group here that’s really worked through the right way to look at this information to say something definitive to the physician. I’m comfortable that there’ll be a state change in diagnostics. Our real challenge is to go from this very short-term research proteomics thing to trying to make a difference.


How valuable is it to be the first to offer a product? What’s the advantage you’ll be able to offer?

It’s clearly better to be first if the product you have is good, and if you don’t turn out to be first, then it’s clear that you minimize the value of that in discussions like this! Obviously you’d like to be first, but in the end what really matters is the scalability of the platform, the limits of detection or how low a protein can be in serum, the accuracy, the [binding] coefficients, the variability, and the [covariance]. What we’re convinced about is that having a photoaptamer covalently attached to the array, and a protein that’s covalently attached to that photoaptamer changes the game. It means washing in a vigorous way that you cannot do in any other kind of capture, and so the backgrounds go down. The limit is always signal to noise. We think being able to wash harshly is a way of lowering the noise, and we think that advantage is enormous. But if you gave me absolutely crappy high noise and on a 47,000-feature chip [right now], I guess I’d rather have that. But that isn’t going to happen.

What is your motivation? Have you always been interested in medicine?

I had a really nice academic science career. I loved teaching graduate students, I loved discovering new stuff, but at the end I always felt a teeny bit guilty about being part of that wonderful enterprise, but never doing anything that was really right-on useful for the people that were paying for it — the taxpayers. I think I have always had this secret desire to be a real doctor rather than this fake-o doctor. I had this chance in the early ‘80s to be a part of Synergen and that just changed my life. I hung out there for seven years and I began to think about medicine, which I had never thought about before. From ‘82 on I always wanted to contribute to healthcare because it just seemed that’s the reason the NIH gave us all the money in our academic labs. The personal motivation now is to actually do something that helps — especially because I’m 60 years old and see all these people getting sick. We discovered a lot of neat stuff during those 30 years [in academia] but none of it was particularly useful to anybody. We learned a lot about how biology works — that’s great. But I’m really driven now to do something short-term useful. Sounds sort of soppy doesn’t it?

The Scan

Sick Newborns Selected for WGS With Automated Pipeline

Researchers successfully prioritized infants with potential Mendelian conditions for whole-genome sequencing or rapid whole-genome sequencing, as they report in Genome Medicine.

Acne-Linked Loci Found Through GWAS Meta-Analysis

Researchers in the European Journal of Human Genetics find new and known acne vulgaris risk loci with a genome-wide association study and meta-analysis, highlighting hair follicle- and metabolic disease-related genes.

Retina Cell Loss Reversed by Prime Editing in Mouse Model of Retinitis Pigmentosa

A team from China turns to prime editing to correct a retinitis pigmentosa-causing mutation in the PDE6b gene in a mouse model of the progressive photoreceptor loss condition in the Journal of Experimental Medicine.

CRISPR Screens Reveal Heart Attack-Linked Gene

Researchers in PLOS Genetics have used CRISPR screens to home in on variants associated with coronary artery disease that affect vascular endothelial function.