Skip to main content
Premium Trial:

Request an Annual Quote

Emerald Logic Applies Evolutionary Computing to Biomarker Discovery


Informatics startup Emerald Logic is hoping to make a name for itself in the biomarker identification market by offering analysis services based on evolutionary computing techniques.

Emerald's CEO Patrick Lilley told BioInform that he and Patrick O'Neill, the firm's chief technology officer, founded the company in 2011 to commercialize analysis software they'd developed based on mathematical principles from evolutionary computing.

He said that they'd initially offered their services to a broad range of vertical markets, including healthcare, finance, defense, and intelligence, but that they now focus almost exclusively on healthcare, targeting pharmaceutical companies and diagnostic developers in particular.

He said that Emerald decided to focus on these markets because it felt that it could offer companies in the space software for biomarker discovery that was faster and more accurate than existing tools.

The company offers services on its proprietary analysis engine, called Fast Collective Evolution Technology, or FACET. The system identifies biomarkers from microarray and protein assays and combines them with other sorts of data such as demographics and medical history in order to develop classifiers to screen patients for diseases and drug response.

Lilley said that the firm charges its customers a flat project fee, which varies depending on the complexity of the analysis that’s required, and it also collects a "percentage of the economic benefit." So, for example, if FACET is used as part of a diagnostic test in a clinic, "we get a percentage of the reimbursements," he explained.

Since its launch, Emerald has worked primarily with clients that are studying neurological diseases. One such client is Irvine Calif.-based neuroinformatics firm Medical Care Corporation. It also has several customers researching autism, Lilley said, but it is now looking for collaborators that can help it move its platform into other disease areas.

Emerald has had some experience outside nervous system disorders, Lilley said. For example, it analyzed data for a drug trial for an undisclosed pharmaceutical customer that looked for biomarkers linked to adverse events such as heart attacks. This year, the firm will focus on wooing customers studying cancer, heart disease, and diabetes, he said.

Lilley said it will do that by forming special collaborations like one it has with King's College London, where it is working with researchers to identify blood markers for Alzheimer's disease and other neurological conditions.

Emerald and KCL researchers worked on two projects last year that focused on identifying blood biomarkers for Alzheimer's and combining this data with other information so that the hospital could use it to screen patients.

Both projects focused on using FACET to find Alzheimer's biomarkers. In one six-week study in late 2012, the partners analyzed more than 11,000 biomarkers from 245 study subjects and identified 14 useful markers.

They then built a mathematical classifier that used the markers they'd found, demographic information, and data from tests for apolipoprotein E — which is implicated in Alzheimer's cases. The classifier was used to distinguish between Alzheimer's cases, controls, and individuals who had mild cognitive impairment.

Lilley said that KCL plans to further research the biomarkers they've discovered and that they'll publish some papers about the Alzhemier studies, with his firm's help, at a later date.

The benefit to Emerald of these kinds of collaborations is that the company gains access to any pharma and diagnostics developers who try to license biomarkers from its partners, he said.

KCL, for example, has "introduced us to several companies … and in those cases, we are pursuing commercial opportunities," he said.

Also, these sorts of partnerships give Emerald a chance to prove its mettle to potential clients in the predictive biomarker market, Lilley said. That should also help it compete with competing platforms offered by firms like GNS Healthcare and newly minted Ayasdi who, according to Lilley, target some of the same customers that Emerald does.

Partnering with KCL, for instance, helped Emerald "build a body of work" and also to validate its technology on multiple datasets, he said.

He said that the company will continue to use this approach when it begins to apply FACET to other disease conditions this year. Emerald already has some collaborations lined up with other organizations, Lilley said, but he declined to disclose who these new partners are.

Toward the end of the year, Emerald plans to release a downloadable version of FACET that customers can purchase. The firm will also increase its headcount from 10 to at least 20 employees by next year, Lilley said.

Emerald plans to will retain some of its non-life science clients, particularly in finance, although the bulk of its business going forward will focus on drug development, Lilley said.

Filed under