Mendelspod Podcast

For years, precision oncology has largely been discussed through the lens of breakthrough drugs. But there’s another story running underneath modern cancer care: the quiet rise of companion diagnostics. These tests are increasingly deciding who receives those therapies in the first place. In many cases, the real bottleneck is no longer discovering a drug target. It’s building a reliable system for identifying the right patient at the right moment in the disease. That challenge sits at the center of this conversation with Rita Shaknovich, Chief Medical Officer for Life Sciences and Diagnostis, and Karina Kulangara, Associate Vice President of R&D in Companion Diagnostics at Agilent Technologies.
Agilent has always had a major role in this field. Rita and Karina explain how companion diagnostics evolved from the original Herceptin test into a vision for a much broader ecosystem spanning pathology, automation, regulation, and global clinical deployment.
We dive into Agilent’s recent FDA approval expanding PD-L1 IHC 22C3 PharmDx into ovarian cancer, a development both guests describe as particularly meaningful given the historically poor outcomes associated with the disease. As Rita puts it: “Precision medicine is based fundamentally on scientific truth . . . it brought real results for patients. It brings better survival for patients, fewer side effects from the medication.”
Karina offers one of the clearest explanations we’ve heard for why immunohistochemistry or IHC has endured so long in modern oncology. “It’s the ability to detect protein biomarker in the spatial context of the tissue,” she explains, emphasizing that location and cellular context can fundamentally shape how therapies work.
What emerges is a picture of precision oncology that is becoming less exotic and more routine. We’re talking not just new drugs, but an entire clinical and technological infrastructure which is designed to match therapies to biology more effectively and over time.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.mendelspod.com/subscribe

Dr. Eric Green returns to Mendelspod in a new role: Chief Medical Officer of Illumina. After more than three decades at the National Human Genome Research Institute, where he helped guide genomics from research initiatives to clinical reality, he now joins one of the industry’s most influential companies at a moment when the field is expanding beyond DNA alone.
Green takes us on a tour around the world of multi-omics, which he says is not a branding exercise but a practical response to the limits of sequence data by itself. Genomics remains foundational, but many clinical questions require additional layers of biology, including RNA, epigenomics, proteomics, and single-cell analysis. As he puts it, “DNA sequence alone may not reveal it.”
The discussion highlights rare disease as one of the clearest examples. Genome sequencing can solve roughly half of suspected cases, Green notes, but many patients remain undiagnosed because the relevant signal may lie in RNA splicing, epigenetic regulation, structural variation, or downstream protein effects. In those settings, multi-omic approaches can provide the missing evidence needed to move from uncertainty to diagnosis.
In oncology, the challenge is different. Cancer genomes can be highly complex and heterogeneous, making it difficult to distinguish driver events from background noise. That is one reason why researchers and clinicians are increasingly incorporating methylation markers, transcriptomic data, and proteomic signals into early detection, disease sub typing, and monitoring strategies.
Green also emphasizes that the next bottleneck may be less about generating data than interpreting it. “The human brain is not going to be the thing that’s going to crack this nut,” he says. “It’s going to be AI and computational biology.”
The result is an overall picture of where the field may be headed as we go from genomic medicine to a broader molecular medicine with multiple data types that will improve diagnosis, stratify disease, and guide care worldwide.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.mendelspod.com/subscribe

This is a free preview of a paid episode. To hear more, visit www.mendelspod.com

For years, proteomics was described as the missing layer of biology. Why missing? Because measuring proteins at scale turned out to be vastly harder than sequencing DNA.
That may finally be changing.
In today’s episode Theral speaks with Dr. Yan Zhang, President of Proteomic Sciences at Thermo Fisher Scientific, about the rapid evolution of large-scale p…

We’ve gotten very good at reading DNA. We’re just beginning to understand how to read its state.
On today’s show, Rob Osborne, Senior Vice President of R&D at Biomodal, discusses new evidence that separating two epigenetic marks—5-methylcytosine and 5-hydroxymethylcytosine—can improve early cancer detection from liquid biopsy. In a recent Nature Communications Medicine study, his team showed that analyzing these signals independently in circulating DNA significantly enhanced detection of Stage I colorectal cancer compared with approaches that combine them.
The advance does not require new sequencing hardware. Biomodal’s approach uses a sample preparation kit compatible with existing platforms, paired with bioinformatics tools, potentially lowering the barrier to adoption while expanding the information content of standard sequencing workflows.
The underlying insight is biological as much as technical. Most methylation assays collapse 5mC and 5hmC into a single signal, masking early transitions in gene regulation. Osborne describes this as “squishing them into one output,” a simplification that can obscure meaningful changes in disease onset and progression.
By separating the signals, the study identified patterns that emerge earlier in tumor development, offering a more sensitive window into disease biology.
But the deeper message of the interview is that this work may only scratch the surface. “I think that we’re just at the beginnings of really understanding this biology,” Osborne says.


This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.mendelspod.com/subscribe

This is a free preview of a paid episode. To hear more, visit www.mendelspod.com

On today’s show, a fascinating discussion about digital twins for medical research— leading to the 64 million dollar question of how much of the current AI wave in healthcare may eventually prove real.
Drawing on her background in cancer research and now as CEO of Concr, Dr. Irina Babina joins us to argue that the future of oncology may depend less on g…