Sensitive and specific biomarkers can help accurately identify complex diseases early as well as provide clinicians with information useful for directing treatment, identifying disease endotypes, stratifying patients for clinical trials and predicting patient outcomes, including prognosis and adverse responses. An ideal biomarker is accessible, easy to measure, specific to the disease, and reproducible. Genomics and proteomics have historically produced valuable biomarkers. Genomics research has uncovered mutations with propensity for disease, such as BRCA1, while proteomics research has identified disease associated proteins, such as Tau protein and beta-amyloid (this issue chapter 4) in Alzheimer’s. Both are helpful in studying disease and determining patient care, but lacking in early detection and prognostic value where the ideal biomarker may considerably improve patient outlook. Recent technological advances in microarray and machine learning technologies have made it possible for scientists to take advantage of patient antibody signatures.
Antibodies are ideal biomarkers because they are manifestations of the actual disease, occurring early, before symptoms, and persisting through the duration of the disease. Antibodies are highly specific, easy to obtain and measure and can represent different aspects of disease, enabling a mechanistic view of disease pathophysiology. New high throughput technologies including protein arrays and machine learning have enabled researchers to identify antibody signatures, or panels of disease associated antibodies, with unprecedented insight. Signatures have much greater diagnostic and prognostic potential than single proteins or gene mutations (Bizzaro, 2007; Kathrikolly et al., 2022). In particular, an individual’s immunosignature is emerging as one of the most sensitive, accurate and predictive sources of disease pathophysiology. Immunoprofiling is not new, but the ability to quantify thousands of antibodies from a small blood sample with protein microarrays provides a new unheralded level of patient detail. A new generation of biomarkers is emerging, capitalizing on the convergence of high throughput measurement techniques, advanced bioinformatics and new machine learning models. In this eBook, we review biomarker discovery, emphasizing immunoprofiling with autoantibody signatures, that once deconvoluted with machine learning, can provide insights into disease progression, potential treatment adverse effects and mechanistic information about disease physiology.