For SEED Grant Semifinalist Divergene, Software Is All In The Genes

Kam To is the Co-Founder of Raleigh-based Divergene. The startup is a semifinalist for one of NC IDEA's $50K SEED grants.

Kam To set out to create a startup that would transform the clinical genomic space through software driven by AI and machine learning.

In 2021, To and co-founder Paola Correa launched the Raleigh-based SaaS startup Divergene to offer genomic data and sequencing services that To said was not available in the market.

After a gene therapy company came to To about issues they were facing with their work, he said he wanted to help find a solution.

“It came out from a need from a customer, and then it evolved,” he said.

Divergene was recently named a semi-finalist for one of NC IDEA’s $50K SEED grants.

To said he has been involved in startups for most of his life; Divergene is the fourth technology-related one he has worked on. Though it’s his background specifically in the clinical genomic space that makes him a perfect fit for this new venture.

He recognized that often, researchers do not have expertise in translating genetic data. Since data is unstructured and not readable by humans, To said many parties needed someone to generate their data into understandable forms, like graphs and dashboards, so it could be analyzed. Recognizing the struggle being faced, To said that he, Correa and their team stepped in solve the problem.

“We process it and then make them interpretable,” he said. “And then we can generate an interpretation to help them better understand the genomic data.”

To said Divergene is a B2B startup and that they market to companies focusing on gene therapy, cell therapy and gene editing. He said their customers also include Contract Development and Manufacturing Organizations and Contract Research Organizations.

Two “pillars” of services

For their customers, To said they offer two primary pillars of services. He said these services are important because the gene of a virus must be sequenced accurately before scientists can know how to attack it. To said that sequencing allows the company to know that the genomic information on the virus is accurate.

“Think of it as a missile,” he said. “You want to have a very precise, guided missile to shoot the target. But if the missile is not guided, or there is no payload, it’s not going to deliver to the right place.”

The first pillar of service that Divergene offers is what To calls the “Off the Shelf Service.” Due to the high cost of sequencing machines, To said that not all companies have the ability to do it themselves. So in this first pillar, the customer will send in a viral vector sample and have it sequenced for them. Once finished, the customer will receive a generated report in an interactive dashboard that outlines what was found. To said this is the most-used pillar.

The second pillar applies to data that has already been sequenced either by the customer or by other sources. In these cases, Divergene processes the raw data, develops a pipeline for customers to analyze and then sends a report back to them.

To said they have also been able to use machine learning to identify virus contamination more accurately.

“Using AI, we were able to increase up to a 30 percent accuracy of identifying contamination within viruses,” he said.

To also said that pricing is dependent upon which pillar is being used. For the first pillar, he said that the cost is related to the number of samples that need to be sequenced. He said the pricing increases with more samples.

The second pillar is priced on a subscription model. After a required initial setup fee, customers can process an unlimited number of samples within their pipeline for an annual rate.

Although Divergene is still in its early stages, To said that they have a dozen customers already, many of whom are recurrent.

“We have already proven the concept and we already generate traction from customers,” he said. “I think it’s a very viable and feasible service and product that we have, especially utilizing the machine learning.”

With Divergene being currently bootstrapped, To said they are not actively looking for funding aside from non-dilutive grants like NC IDEA’s. But To said he is willing to have discussions with interested investors for the future.