Genomics, A.I. Calculate Immune Health

(Blue Diamond Gallery)

12 Mar. 2019. An international team designed a technique to gauge a person’s immune system age and health, by analyzing genomic and cellular indicators with machine-learning algorithms. Researchers from Stanford University in California, along with Technion Institute and the start-up company CytoReason in Israel describe their technology in the 6 March issue of the journal Nature Medicine (paid subscription required).

Researchers led by immunologists Mark Davis at Stanford and Shai Shen-Orr at Technion are seeking more accurate and reliable measures of immune system health, which declines with age. Yet, while new therapies are developed that harness the immune system to treat diseases like cancer and disorders from malfunctioning immune systems, clinicians still lack a reliable metric or indicator for characterizing this key function in the body.

“It’s been sixty years since the last immunological benchmarks, complete blood counts, were introduced into general medical practice,” says Davis in a CytoReason statement. “This much more sophisticated method reflects the tremendous explosion of knowledge generated in the field.”

The Stanford-Technion team tracked 135 healthy individuals for 9 years, taking periodic measures of various immune system indicators. These measures include genes expressed in whole blood samples, responses of cells to cytokines or signaling enzymes emitted by immune system cells, and characteristic traits of specific cell subsets in the body. The huge store of data collected from these volunteers over time provided raw material for computational techniques, including machine-learning algorithms and natural language processing, developed in Shen-Orr’s lab at Technion in Haifa called a cell-centered analysis model.

These algorithms and model enable the researchers to compute a composite indicator for immune system health over time that the research team calls immune age. “The immune age is a biological clock,” notes Shen-Orr, “that will help to identify in individuals, the decline and progress in immunity that occurs in old age, with the aim of determining preventive measures and developing new treatment modalities to minimize chronic disease and death.”

The team validated the immune age model with 2,000 patient records in the Framingham Heart Study, a continuous tracking of cardiovascular health with thousands of adults in Framingham, Massachusetts since 1948. The researchers report the immune age model predicted mortality from any cause among those selected beyond the risk factors routinely identified in the Framingham study.

CytoReason is a 3 year-old enterprise in Tel Aviv that applies Shen-Orr’s research to drug discovery and development. The company uses computational methods to identify new disease targets, provide insights into mechanisms of diseases and drugs working in the body, and explain differences in response interactions between drugs and diseases. CytoReason collaborates with pharma and biotech companies to help guide their drug discovery decision-making, but is not developing its own therapy pipeline.

The company says immune age measures can help plan clinical trials, particularly studies of drugs invoking the immune system, by supplementing chronological age as a criterion for inclusion. The CytoReason web site provides case studies of its technology with treatments for inflammatory bowel disease, atopic dermatitis or eczema, and melanoma, an aggressive form of skin cancer.

More from Science & Enterprise:

• Digital Tech, A.I. Assessed for Heart Valve Problems
• High-Speed Screens, A.I., Boost Nanoscale Medicines
• Large Molecule Library Enables On-Demand Drug Discovery
• Algorithms Identify More Genetic Syndromes
• Tissue Chip, A.I. Companies Partner on Heart Drug Discovery

*     *     *