- Funding to Date*
A huge problem in genetics is understanding what a particular mutation in a gene will do if it has not been observed before. BRCA1 is a well known mutation that causes susceptibility to breast cancer. We know that at the 185th nucleotide a deleted Adenine or Guanine can predispose one to breast cancer. This is because after sequencing thousands of breast cancer genes we observe this deletion event over and over and therefore have correlated the two events.
If if you sequenced your gene for breast cancer and they found a sequence that wasn’t normal but also not seen as cancer causing, what would you do? That sequence is called a Variant of Unknown Significance, or as is said in the industry, a VUS. While doing their PhDs and postdoctoral work at Stanford, Carlos and Alex had developed an approach to solve the mystery of VUSs. They used machine learning to look at the particular variant and predict its effect on the protein and the level of harmfulness of the variant. Later on, they developed a lab based assay to corroborate the interpretations, and further train the network (a critical piece of the puzzle we call “wet-in-the-loop”).
The largest genetic testing service to individuals and healthcare providers, Invitae, needed a solution to the VUS interpretation problem. With his skillset in hand, Carlos approached and closed, within four months, a pilot with Invitae to interpret VUSs in their samples. The pilot went so well that Invitae offered to acquire Jungla to incorporate their technology into Invitae’s offering.