Search for: All records

ABSTRACT The physical organization of DNA within the nucleus is fundamental to a wide range of biological processes. The experimental investigation of the structure of genomic DNA remains challenging due to its large size and hierarchical arrangement. These challenges present considerable opportunities for combined experimental and modeling approaches. Physics‐based computational models, in particular, have emerged as essential tools for probing chromatin structure and dynamics across a wide range of length scales. Such models must necessarily be capable of bridging scales, and each scale presents its own subtleties and intricacies. This review discusses recent methodological advances in genomic structural modeling, emphasizing the need for multiscale integration to capture the hierarchical organization and molecular mechanisms that underlie chromatin structure and function. We present an analysis of state‐of‐the‐art methods, as well as a perspective on challenges and future opportunities across length scales ranging from bare DNA to nucleosomes and chromatin fibers, up to TAD and chromosome‐scale models. We emphasize models that connect genome organization to gene expression, models that leverage emerging machine learning capabilities, and models that develop multiscale approaches. We examine gaps in experimental data that computational models are poised to address and propose directions for future research that bridge theory and experiment in DNA structural biology. 

Hydrodynamic theories effectively describe many-body systems out of equilibrium in terms of a few macroscopic parameters. However, such parameters are difficult to determine from microscopic information. Seldom is this challenge more apparent than in active matter, where the hydrodynamic parameters are in fact fields that encode the distribution of energy-injecting microscopic components. Here, we use active nematics to demonstrate that neural networks can map out the spatiotemporal variation of multiple hydrodynamic parameters and forecast the chaotic dynamics of these systems. We analyze biofilament/molecular-motor experiments with microtubule/kinesin and actin/myosin complexes as computer vision problems. Our algorithms can determine how activity and elastic moduli change as a function of space and time, as well as adenosine triphosphate (ATP) or motor concentration. The only input needed is the orientation of the biofilaments and not the coupled velocity field which is harder to access in experiments. We can also forecast the evolution of these chaotic many-body systems solely from image sequences of their past using a combination of autoencoders and recurrent neural networks with residual architecture. In realistic experimental setups for which the initial conditions are not perfectly known, our physics-inspired machine-learning algorithms can surpass deterministic simulations. Our study paves the way for artificial-intelligence characterization and control of coupled chaotic fields in diverse physical and biological systems, even in the absence of knowledge of the underlying dynamics.