1 March 2017. Freenome, a company applying machine learning to analysis of liquid biopsies, is raising $65 million in its first round of venture financing. The funding round for the 3 year-old enterprise in South San Francisco, California, is led by technology investment company Andreessen Horowitz.
Freenome’s founders, veterans of other health care and life science companies, began the enterprise to improve the state of cancer diagnostics, making it possible to discover cancer earlier in the development of the disease, when individuals have more treatment and even prevention options. The company uses the term “freenome” as well to describe the cell-free genetic material floating in the blood stream, which makes up some of the raw material it analyzes for cancer diagnostics.
The problem faced by many current cancer screening methods, such as prostate-specific antigens or PSAs for prostate cancer and mammograms, says Freenome, is their imprecision that leads to erroneous results, such as false-positives, resulting in unnecessary procedures such as biopsies, and even treatment. Part of the problem with current screening methods, says the company, is the application of simplified models to understand a dynamic, complex disease such as cancer. Moreover, physicians need to know more information about the particular disease than just its presence or absence in the patient.
The technology developed by Freenome combines genomics, computational biology, and artificial intelligence to analyze simple blood samples for this dynamic, complex environment. Rather than focusing on a few genetic mutations suspected of causing a person’s cancer, Freenome’s process analyzes the full gamut of molecular clues in the blood, looking for evidence of immune-system or metabolic changes, as well as DNA from cells emitted by a tumor.
Freenome’s technology is built on its adaptive genomics engine that applies deep learning, an artificial intelligence and machine learning process, to discern underlying patterns in relationships, and build those relationships into knowledge bases found today in a number of disciplines. In this case, the adaptive genomics engine seeks to identify unique molecular signatures that not only detect the presence of cancer, but also provide a detailed picture of the disorder for the cancer specialist to prescribe a precise course of treatment. By making the test simple and accurate, says the company, many cancers could be caught earlier and treated successfully.
Freenome’s co-founder Gabriel Otte believes the technology may even have enough power and precision to find indicators of potential cancer early enough in the process to take steps to prevent the disease from occurring. “One day,” says Otte in a company blog post, “not only will we be able to find diseases BEFORE they happen, but we may be able to suggest next steps that will empower patients to avoid those diseases on their own.”
Otte says Freenome is working with researchers at University of California in San Francisco and San Diego, as well as Massachusetts General Hospital to further refine the technology. “We are also working with five pharmaceutical companies,” Otte adds, “to give an accurate indication of the treatments that a patient would most likely to respond to once we’ve detected the cancer.”
Andreessen Horowitz, a seed-round investor, is leading Freenome’s venture round, joined by seed investors Data Collective and Founders Fund. Joining in the round are new investors Google Ventures, Polaris Partners, Innovation Endeavors, Asset Management Ventures, Charles River Ventures, and Spectrum 28. Freenome raised $5.6 million in its seed round in June 2016.
More from Science & Enterprise:
- Test of Nasal Skin Cells Found to Detect Lung Cancer
- Alliance to Study Brain Tumor Precision Diagnostics
- Illumina Partnering with IBM, Phillips on Genomics Tools
- Trial to Test Liquid Biopsy for Breast Cancer Complication
- NantHealth, TransMed Partner on Cancer Genomics
* * *
You must be logged in to post a comment.