Seven years ago, scientists at Illumina, the DNA sequencing company in San Diego, were running a study seeking DNA fragments in pregnant people’s blood that might suggest chromosomal abnormalities.

During the research, a pathologist discovered something unexpected in 10 of the blood samples.

Instead of chromosomal disorders, the testing showed DNA abnormalities.

This didn’t make sense to the researchers. But there was some suspicion that cancer was involved.

Investigating further, the researchers learned that one of the 10 participants had in fact received a cancer diagnosis, and others in that subset also had cancer — despite the fact that they had no symptoms.

This convinced Illumina’s leadership to create a spin-off company in 2016 called GRAIL that combined advances in human genomics with machine learning data science.

That resulted in the development of a new blood test from GRAIL called Galleri, which can detect early stages of cancer before a person has symptoms.

In clinical studies, an earlier version of Galleri showed the ability to detect more than 50 types of cancer — 45 of which lack recommended screening — through a single blood draw.

A study done this year with 6,600 participants that included the Mayo Clinic discovered 29 signals that were followed by a cancer diagnosis. Another study reported a false positive rate of less than 1 percent.

Overall, GRAIL has enrolled more than 134,000 people in clinical trials of the blood test. The test also shows where the cancer is located.

“When a cancer signal is detected in these trials, the test also pinpoints where the cancer is located in the body with high accuracy,” Dr. Joshua Ofman, chief medical officer and head of external affairs at GRAIL, told Healthline. “This helps healthcare providers to determine next steps for diagnosis and care.”