de Pablo Group

Liquid crystals (LCs) are a phase of matter that flows like a liquid, but the orientations of the molecules are highly ordered over a very long range. This presence of long-range orientation results in interesting behavior of systems that employ LCs. In our group, we model LCs on multiple scales in an effort to engineer new applications for the laboratory and industry. At the atomistic level, we investigate the behavior of LCs near surfaces to determine the types and strength of anchoring present at different surfaces. At a mesoscale, we study systems mixtures and determine the accessibility different phases of LCs. On the largest scales, we investigate the behavior of particles, from the nanometer to micron scale, and observe their behavior in an LC solvent; the presence of defects in the LC has a marked effect on particle behavior, so by controlling the defect with fields (flow, electric, magnetic,etc.), we can dictate particle behavior in a well controlled manner.  We also study how introducing active biological agents, such as bacteria or myosin motors, influences the dynamics of LC systems.

Our research group uses coarse grain models to study the biophysics of DNA and chromatin. Recent efforts have included:

1. Developing hierarchical coarse-grained models to study DNA and chromatin at multiple length scales.
2. Relating nucleosome interactions to chromatin fiber structure
3. Exploring collective motions in chromatin using nonlinear manifold learning
4. Examining the effect of DNA-protein interactions on the structure and dynamics of chromatin

With the help of modern polymer chemistry, macromolecules can now be synthesized with exquisite control over architecture, composition, and monomer sequence. These parameters work cooperatively to control the dynamics, self-assembly, and morphology of the resulting materials and determine how they may be applied in new technologies.  To understand and predict these properties, we develop and utilize multi-scale Monte Carlo and molecular dynamics simulations as well as novel theoretical treatments.  Our current ares of research include:

• Controlling complex coacervation of polyelectrolytes with random charge sequences
• Structure and rheology of "miktoarm" polymers and their blends
• Using multi-block co-polymers to design complex nanostructures in solution
• Dynamics of interlocking polymers

Molecular simulations are typically limited by the time scale of sampling. With reasonable amount of computational resources one can only simulate on the order of hundreds of nanoseconds. On the other hand in real complex systems most of the phenomena of interest occur at orders-of-magnitude longer time scales. To solve such problems, we develop new advanced sampling algorithms that can accelerate discovery in these systems in both Monte Carlo (MC) and Molecular Dynamics (MD) simulations, which have been incorporated into the Software Suite for Advanced General Ensemble Simulations (SSAGES) code.

Many soft matter systems exhibit a wide range of length scales, making them challenging to simulate and study.  For example, colloidal suspensions are naturally represented as discrete particles embedded in a continuum since the colloids are much larger than the individual molecules of the solvent.  To couple the dynamics of particles and continua, we have developed the Continuum-Particle Simulation Software (COPSS) code, which uses innovative numerical methods to efficiently simulate mesoscopic soft matter systems.

The SSAGES and COPSS codes are available at: http://miccomcodes.org/

Machine learning provides a powerful set of methods for extracting useful information from large and complex data sets, such as those routinely generated by molecular simulations.  Our group uses these two complementary tools to gain insight into the properties of molecular systems and to make predictions for materials design.  Recent efforts include:

• Using artificial neural networks (ANNs) for force-biasing during molecular dynamics simulations
• Predicting electronic structure properties of conjugated polymers at coarse-grained resolution using ANNs
• Inverse design of materials based on desired physical properties

Metamaterials (literally going-beyond-materials) are a paradigm of engineered materials that possess properties beyond what is conventionally observed. In these projects, we apply molecular simulation techniques to model systems comprising of disordered networked materials which can then be optimized for a target stimulus response. These efforts lie at an exciting intersection of biology, physics, computing, and materials science.

In the past we have designed materials with a target response that is mechanical in nature (auxetic metamaterials), and at present we are exploring bioinspired materials in the context of allostery. Allosteric proteins can alter their active site when an effector is bound to a physically distant allosteric site, regulating their biological function. We model a protein as a network of interacting residues, the properties of which may be altered to achieve an allosteric response. These studies can lead to useful insights about biological regulation of protein function, and the lessons learnt can then be transferred to creation of synthetic materials with bio-inspired properties.