Toward physical principles in biology : mechanical fingerprints of biomolecules, remote chromosomal interactions, and viral invasion strategy

This dissertation develops quantitative approaches to understanding biological phenomena -- arguably the most complex ones around -- with the language and tools of theoretical physics. We attempt to establish the general principles that unify diverse phenomena of the living world, and to capture some of these principles in the form of statistical-mechanical theories. The emphases are put on the analytical (rather than purely numerical) nature of the theories and on their power to generate experimentally testable predictions. The developed theories, being rooted in physics, address a number of biologically significant questions : (1) How can we decode the mechanical fingerprints, or the response to force, of biological macromolecules? (2) How do remote DNA segments find each other in the crowded environment of the cell nucleus to establish genomic interactions? (3) How do viruses infect cells on remarkably short time scales despite high energy barriers? To address the first question, we establish a model-free transformation that decodes the mechanical fingerprints of biomolecular interactions probed in force spectroscopy experiments. The transformation converts the mechanical fingerprints into a form that reveals the activation barriers and timescales of a biological process. As a special but important case, conformational dynamics that proceed via an intermediate state are investigated and captured in the form of a predictive analytical framework. To address the second question, we develop a theoretical approach to V(D)J recombination -- the genetic mechanism that allows the human immune system to respond to millions of different antigens. The theory, applied to 3D trajectories of chromosomal segments in living B lymphocytes, establishes the mechanism by which chromosomal DNA moves around in the cell nucleus. The established mechanism -- fractional Langevin motion -- allows us to predict the first-passage times for genomic interactions. To address the third question