Cotton Candy and Multidimensional Biomarker Discovery Lead to an Exciting Work Day
Jeff Hiken, R&D Lab Director, and Kevin Flanagan, Scientist, at Cofactor Genomics share how they are using multivariate analyses to identify new multidimensional biomarkers. You can also watch the videos here.
Q: What’s your favorite part about working at Cofactor?
Jeff: Here at Cofactor, we have a weekly Friday talk. Usually the talks are about science or biotech, but everyone in the company participates, and any topic is fair game. We have a group of people with wide ranging backgrounds and interests, and we’ve had talks on topics from ranging Filipino history, to solar eclipses, to the history of cotton candy…
Kevin: I especially enjoyed the cotton candy taste test. And I agree, whether it’s during the Friday talks or just in day-to-day interactions, I love working with and learning from smart, passionate people with many different backgrounds. Having this great team is great for helping us advance precision medicine, but it also just makes for a really interesting work environment where I constantly learn new things.
Q: Why do you believe multidimensional biomarkers are an improvement over single-analyte approaches? And, What does Predictive Immune Modeling mean to you?
Jeff: Checkpoint inhibitors target immune inhibitory receptors on T cells, and in doing so they can kick-start immune responses in cancer patients. Many patients have shown a dramatic benefit from treatment with checkpoint inhibitors, but unfortunately, it turns out that those patients are in the minority. For example, only about 15 – 30% of patients respond to blockade of the inhibitory receptor PD-1, or its ligand PD-L1.
So, there’s a growing need to develop biomarkers that can help identify which patients will benefit from immunotherapy.
Currently, single analyte assays are used for this, for example measuring PD-1 ligand; or measuring the level of tumor infiltrating lymphocytes.
But single analyte assays often can’t capture the complexity of the tumor microenvironment, and these single analyte assays are proving to be unreliable predictors of clinical outcomes.
At Cofactor, we use multivariate analyses to identify new biomarkers.
Our computational scientists have developed a new discipline that we call Predictive Immune Modeling. With Predictive Immune Modeling, we combine data on immune cell composition of the tumor, RNA expression of important genes that characterize the tumor microenvironment, and if available, clinical outcome data. With machine learning, these multivariate inputs can synergize to define a biomarker with significantly greater predictive value than a single analyte assay.
Q: Why do you believe interrogating RNA is the best way to understand disease?
Kevin: Looking at RNA really is a great approach to finding multi-dimensional biomarkers and driving Predictive Immune Modeling. DNA is pretty much static—it’s the essentially the same in every cell and doesn’t change much throughout your life. RNA, on the other hand, is dynamic. Each cell has unique patterns of expression, and they change in response to development, disease, drug treatments, and more. Because RNA can be analyzed in a high-throughput, quantitative way, we can quickly learn a ton about a disease and the immune system’s response to it, without the need for large samples or qualitative judgment calls.
Questions about Cofactor or our product offerings? Reach out to schedule a time to speak with one of our Project Scientists today.