The human immune system is a complex network, consisting of trillions of cells diligently circulating throughout the body. This intricate web interacts with every organ and tissue, executing a multitude of functions that scientists are still striving to decipher. Due to this complexity, predicting patient responses to treatments can be challenging, leading to significant barriers in drug development and clinical trials, even for drugs that show potential benefits for certain patients.
This challenge often causes pharmaceutical companies to abandon promising drugs due to the uncertainty surrounding patient responses and potential side effects.
Immunai is addressing this issue by creating a detailed map of the immune system. The company has developed a comprehensive database known as AMICA, which amalgamates gene and protein expression data from various cells alongside clinical trial information. This innovative tool is designed to match the appropriate drugs to the right patients.
“Our initial goal was to create what I refer to as the ‘Google Maps for the immune system,’” explains Immunai co-founder and CEO Noam Solomon. “We began with single-cell RNA sequencing and progressively incorporated a range of ‘omics’ data — including genomics, proteomics, and epigenomics — to assess the cellular expression and overall function of the immune system. Our collaboration with pharmaceutical companies and hospitals allows us to thoroughly profile patients’ immune systems during treatments, which helps uncover the root mechanisms of therapeutic action and resistance.”
Immunai’s robust data foundation stems from the unique backgrounds of its founders. Solomon and co-founder Luis Voloch, both of whom possess degrees in mathematics and computer science, designed this platform to merge computational prowess with life sciences. At the time of Immunai’s inception, Solomon was a postdoctoral researcher in MIT’s Department of Mathematics.
Solomon envisions Immunai’s mission as bridging the gap between computer science and life sciences, which have diverged over the years. He highlights the paradox where advances in computing, driven by Moore’s Law, stand in stark contrast to the escalating costs in drug development, which reportedly double every nine years — a trend referred to as Eroom’s Law.
“Why should pharmaceutical companies continue to invest in discovery if it doesn’t yield a return?” Solomon asks. With only a 5 to 10 percent success rate for clinical trials, he sees Immunai’s mapping of the immune system as a solution to enhance the early stages of drug development.
A Change in Direction
Solomon embarked on his educational journey at Tel Aviv University at the age of 14 and completed his computer science bachelor’s degree by 19. He attained two PhDs, one in computer science and the other in mathematics, before joining MIT in 2017 for postdoctoral research.
During his tenure at MIT, he crossed paths with Voloch, who had completed his bachelor’s and master’s studies in math and computer science at the same institution. Their journey took a transformative turn when Voloch’s grandfather underwent cancer treatment that temporarily sent his condition into remission, albeit with severe side effects that forced him to stop medication.
This personal experience motivated Voloch and Solomon to explore how their expertise could assist patients like Voloch’s grandfather. “Realizing our potential for impact was a pivotal moment that led us to pivot from academia to founding Immunai,” Solomon reflects. “That was our motivation.”
They later joined forces with Immunai’s scientific co-founders, Ansu Satpathy from Stanford University and Danny Wells from the Parker Institute for Cancer Immunotherapy. Satpathy and Wells had demonstrated that single-cell RNA sequencing could provide insights into why individual patients have varied responses to common cancer therapies.
The team began examining publicly available single-cell RNA sequencing data and working to correlate specific biomarkers with patient outcomes. Integrating data from the UK Biobank, they enhanced their predictive models. Soon, they were incorporating data from diverse sources, including hospitals, academic institutions, and pharmaceutical companies, to analyze the biological landscape of immune activity utilizing multiomics approaches.
“Single-cell sequencing allows us to obtain metrics from thousands of cells, enabling us to evaluate 20,000 different genes. These metrics help us construct an immune profile,” Solomon explains. “By continuously measuring and comparing pre- and post-treatment data between responding and non-responding patients, we can leverage machine learning to decode the underlying reasons for these variations.”
Collectively, this data and these models form AMICA, which Immunai touts as the world’s largest database focused on immune cells. AMICA stands for Annotated Multiomic Immune Cell Atlas and encapsulates single-cell multiomic data from nearly 10,000 patients, alongside bulk-RNA data from 100,000 individuals spanning over 800 cell types and 500 diseases.
At the heart of Immunai’s strategy lies a focus on the immune system — a complex domain that many other companies tend to avoid.
“We differentiate ourselves from those primarily focused on tumor microenvironments,” Solomon emphasizes. “The immune system is central to all diseases; it governs your body’s response to everything, from viral and bacterial infections to the drugs you take and even the process of aging.”
Transforming Data into Enhanced Treatments
Immunai has established partnerships with some of the world’s largest pharmaceutical firms, assisting them in pinpointing effective treatments and laying the groundwork for successful clinical trials. Their insights guide critical decisions regarding treatment schedules, dosages, drug combinations, and patient selection, among other factors.
“While AI is a hot topic, the most thrilling aspect of our platform is its vertical integration, encompassing everything from wet lab experiments to computational modeling with iterative feedback loops,” Solomon remarks. “For instance, we might conduct single-cell immune profiling of patient samples, upload this data to the cloud, and use our computational models to generate insights. Subsequent in vitro or in vivo validations further refine and enhance these models.”
Ultimately, Immunai aspires to create a future where laboratory experiments seamlessly translate into transformative treatment recommendations that genuinely benefit patients.
“Scientists can effectively cure nearly every type of cancer in animal models,” Solomon observes. “However, there remains a significant gap in translating those successes to human applications. To bridge that gap, many are seeking better ex vivo or in vivo models, but our strategy focuses on creating a more agnostic approach to model systems, feeding our algorithms with expansive data from multiple model systems. We aim to show that our algorithms consistently outperform leading benchmarks in identifying crucial preclinical immune features linked to patient outcomes.”
Photo credit & article inspired by: Massachusetts Institute of Technology
