A Decade of Advances in Single-Cell Functional Immunomics

Since the inaugural GEN report on the subject in 2015, the realm of immunology has undergone a profound transformation. Researchers have grown increasingly proficient at scrutinizing immune cells through the lens of single-cell technologies, yielding a more nuanced understanding of immune-cell diversity. However, a critical gap persisted: molecular examinations often failed to anticipate the actual cellular behavior that would manifest over time.

Immune cells displaying activated molecular profiles would sometimes falter in maintaining enduring antitumor activity, while cells that proliferated efficiently during cultivation did not necessarily ensure persistence or therapeutic efficacy. The intricate biological mechanisms governing therapeutic outcomes unfold dynamically, governed by factors such as behavior, adaptation, persistence, and dysfunction over time.

This necessitated the development of methodologies capable of observing immune cell behavior directly alongside molecular state.

At Northeastern University, our research team concentrated on engineering technologies designed to investigate immune function in a more dynamic and biologically relevant manner. Central to our efforts were the creation of single-cell systems capable of observing immune-cell behavior in real time while retaining the essential biological context required for meaningful downstream molecular analysis.

Over the subsequent decade, breakthroughs in microfluidics, live-cell imaging, computational analysis, and single-cell biology facilitated the direct correlation between observed immune behavior and downstream molecular and clinical data at unprecedented scales. This initial endeavor to accurately observe immune-cell behavior evolved into integrated functional-immunomics systems with the capacity for translational analysis, therapeutic characterization, and sophisticated computational modeling.

These contributions paved the way for the emergence of functional immunomics—an approach that centers on comprehending immune behavior and its relationship to disease and therapeutic outcomes.

A fundamental shift in immunology over the past decade has been the acknowledgment that immune-cell behavior assumes a pivotal role as a biological variable. While traditional molecular approaches continue to offer essential insights into gene expression, signaling states, and cellular composition, functional approaches augment these modalities by incorporating behavioral context into molecular information.

Behaviors, such as persistence, resistance to dysfunction, therapeutic durability, and effective immune responses, are inherently functional properties. Phenomena like serial killing, sustained cytotoxicity, and resistance to exhaustion cannot be entirely elucidated through static measurements alone, as these phenomena emerge over time.

Breakthroughs in microfluidics, live-cell imaging, and single-cell analysis have rendered it increasingly feasible to observe these processes directly. In our research, controlled single-cell pairing systems allowed us to place immune cells into defined microenvironments with tumor targets, enabling the observation of functional behavior prior to downstream molecular analysis.

One of the most striking observations from our early single-cell experiments was the frequent occurrence of immune cells exhibiting robust activation signatures that failed to sustain functional killing over time, while cells that appeared less remarkable on a molecular level often turned out to be the ones orchestrating effective cytotoxic responses.

Witnessing this pattern repeatedly across experiments gradually shifted my perspective from a focus on the molecular composition of immune cells to their actual behavior. This observation gradually prompted a broader shift in the field, with a growing number of academic and industry entities embracing the emphasis on function-linked immune analysis.

My research at Northeastern University laid the intellectual groundwork for this conceptual framework, which was subsequently advanced through Feromics, a functional immunomics company focused on AI-enabled immune analysis. The overarching objective has been to establish function as a fundamental framework for immune analysis, therapeutic development, and predictive modeling.

The advent of artificial intelligence in biology has further accentuated the significance of this transition. Machine-learning systems are inherently influenced by the biological quality and structure of the data utilized for training. Traditional bulk population-averaged datasets introduce biological noise by amalgamating disparate cell states into a single average, whereas function-linked analyses effectively resolve this heterogeneity, yielding cleaner, more therapeutically pertinent immune signatures.

A mixture of cells in varying states—some highly cytotoxic, others exhausted, and still others transitioning between states—may generate a population-level signal that fails to accurately represent any of those states. In practice, this implies that cells displaying canonical activation markers may still exhibit diminished sustained cytotoxic activity when subjected to functional analysis over time.

Functional analysis enables the differentiation of these behaviors, sidestepping the averaging of disparate measurements into a single composite metric. As functional single-cell datasets mature, they facilitate the development of computational systems capable of directly linking immune behavior with therapeutic outcomes.