The Cancer R&D Conference recently held in Newton, MA, was a great opportunity to dive into new ideas and innovation. I couldn’t make it to the whole event (thanks, rhinovirus—I didn’t want to be Typhoid Eleanor), but the sessions I did catch offered some real food for thought. Here are a few that caught my attention.
X. Shirley Liu, Co-Founder & CEO of GV20 Therapeutics, shared how her team uses TRUST to mine for antibody sequences hiding in tumor RNA-Seq data. These are sequences that usually get discarded because they don’t map to a reference genome. But instead of trashing them, Shirley’s team found they might bind to tumor antigens, making them possible drug candidates or biomarkers. It’s a great example of getting more from data we already have.
My former PhD co-advisor, John Quackenbush from the Harvard T.H. Chan School of Public Health, gave a great talk on how to map regulatory networks from transcriptomic data using tools such as LIONESS and PHOENIX. These tools are especially handy for mechanism-of-action studies—when you’ve got transcriptomic data and are wondering, “Ok, now what?” Methods like these can help turn a useless list of genes into a hypothesis that can be tested in the lab.
Sangeeta N Bhatia from the Koch Institute for Integrative Cancer Research, MIT, discussed her work using nanotechnology for cancer diagnostics. One example involved a 100nm particle that binds to a cancer-specific protease and releases a tiny marker (<5nm) that the kidney filters into the urine. This could have all kinds of applications, from paper-based diagnostics to aerosol delivery systems. It’s a simple yet powerful approach to detecting disease.
Ying Xu from Southern University of Science and Technology shared a completely different perspective on cancer. His “stressor-first” paradigm suggests that metabolic changes, such as shifts in pH or ion concentrations, drive cancer, rather than genetic mutations. And it is these stressors that lead to the genetic and epigenetic changes associated with cancer. I think this is a controversial idea, but it could lead to new ways of understanding and treating cancer. I’ve added his papers to my (already overflowing) reading list.
Hui Shen from the Van Andel Institute gave a talk on single-cell methods in ovarian cancer. She started with gene-centric insights from TCGA before shifting to cell states—like how the menstrual phase during which a cell becomes cancerous influences its subtype. Wild, right? Her lab uses tools like STORM-seq to study rare ovarian cancer stem-like cells. This protocol extracts total RNA, uses random hexamer priming, and provides unmatched paired-end sequencing. The results? Full gene coverage, minimal bias, high mapping rates, and 11,000 genes per cell at high read depth. Takara Bio even commercializes it. They’ve identified distinct cell populations and uncovered how differences in cell type and cell cycle impact normalization. This is a great example of how single-cell methods can unravel the complexity of ovarian cancer.
For me, conferences like this are about more than just hearing cool talks. They’re a chance to bring new ideas and tools back to my team at Diamond Age and think about how we can put them into practice for our clients. If you’re grappling with a tricky data problem or just want to explore new ways of looking at your research, let’s talk. We’d love to help you get more out of your data!