Please introduce yourself and tell me a bit about your research.

My name is Itay Tirosh and I'm a professor at the Weizmann Institute of Science in Israel. Our lab investigates the diversity of cells within tumors – how this variation arises, what drives it, and its significance in cancer progression. We explore whether specific cell populations are linked to key tumor characteristics, such as drug resistance, metastasis, or other features.

Traditionally, it has been thought that analyzing the global features of a tumor, such as its driver mutations, would be the key to understanding its phenotypes. However, we are increasingly realizing that phenotypes can emerge from small subpopulations of cells, each contributing in significant ways. A tumor functions as a complex ecosystem, and to truly understand its behavior, including emergent properties, you need to consider all of its components. This is the central focus of our lab: to study this diversity and uncover how it shapes tumor biology.

Can you explain a bit about your approach?

Our approach is somewhat computational: many of us in the lab come from computational backgrounds, which naturally shapes our research. We tend to work on projects that generate vast amounts of data to characterize the tumor ecosystem, followed by extensive computational analysis to make sense of it all. Over the past decade, our primary method has been single-cell RNA sequencing. In collaboration with clinicians, we obtain tumor samples immediately after surgery, dissociate them, and profile them using single-cell RNA Seq. This serves as the foundation for deeper analyses, helping us uncover the diverse cell populations that coexist within a tumor.

One of the key questions that emerged was how to explain the diversity we see within a single tumor. Even when focusing solely on cancer cells, excluding the microenvironment, we still observe significant variation among them. Over time, it became clear that genetic mutations alone are not sufficient to account for this diversity.

We realized that tumor cell diversity must be influenced by factors beyond genetics – likely tied to the location of cells within the tumor. Over the past few years, we’ve shifted from simply mapping cells without considering their spatial context to actively analyzing their positions. Our focus is now on understanding how a cell’s identity and state are shaped by its environment, interactions with neighboring cells, and the overall organization of the tumor. Much of our recent work has been in spatial mapping.

What excites you about the field and how do you see it progressing in the next five to 10 years?

Single-cell RNA sequencing has been our primary tool – not just within our lab, but across the broader field of omics cancer research over the past decade. However, we are now seeing a gradual shift toward spatial methods, which I believe will become increasingly dominant in tumor analysis over the next few years. As researchers strive to extract the most comprehensive information possible, spatial analysis is poised to replace single-cell RNA sequencing as the preferred approach. Multiple technologies are emerging, each with unique advantages, and as they continue to evolve, they will fundamentally change not only how we analyze tumors but also how we understand the tumor ecosystem itself.

Tell us about your collaboration with Abcam

Initially, we began using an antibody panel. Our spatial approach incorporates both transcriptomics and proteomics, each offering distinct advantages. Proteomics, in particular, allows us to analyze proteins directly, which are more functionally relevant, and it provides high-quality, true single-cell resolution – something that transcriptomics struggles with, as separating individual cells can be challenging.

“For spatial proteomics, we needed a robust antibody panel, which led us to collaborate with Abcam.”

Together, we developed a panel not only for conventional targets, such as immune cell types and states, but also for novel cell states identified through single-cell sequencing – states that previously lacked well-suited antibodies for precise identification.

We developed this antibody panel using Abcam to identify all the cell states we discovered through single-cell analysis. More recently, we’ve expanded its scope beyond cell states to also target the extracellular matrix (ECM), an often-overlooked component of the tumor. Recognizing that a complete understanding of tumor behavior requires more than just analyzing the cells themselves, we are now focusing on the ECM’s role in shaping cell identity, interaction potential, and other key characteristics.

“A big part of our lab’s work relies on antibody panels, and many of the best antibodies we know come from Abcam. To ensure high-quality antibodies for all our target molecules, Abcam was the most obvious choice. Their extensive catalog and reliability make them a great fit for our research.”

What groundbreaking developments are shaping cancer research, particularly in proteomics?

There are so many groundbreaking developments at AACR – it’s hard to pick just one. Immunotherapies continue to advance rapidly, and KRAS inhibitors are making significant strides. The field is evolving at an incredible pace, and these innovations could truly transform how we stratify and treat cancer patients.

In terms of emerging proteomics technologies, we’ve been using Codex, also known as Phenocycler, for our antibody panels, and we find it highly efficient. We expect it to become more widely used in the near future. One advantage that doesn’t get enough attention is its large capture area, which allows for tissue microarray analysis – enabling us to study many tumors simultaneously. Some other technologies have much smaller capture areas, making this an important distinction.

How do you see AI impacting cancer research?

Data analysis in this field is incredibly complex, with many challenges to overcome. AI is gradually revolutionizing analytical methods, and we see it increasingly integrated into research workflows. Of course, AI is infiltrating nearly every discipline, but in terms of our own work, we haven’t been among the earliest adopters.