Abstract: Beginning with the famous 1929 statement of Paul Dirac concerning the complexity of the application of the fundamental equations of quantum mechanics to chemistry, computers and quantum chemistry have been joined at the hip. Indeed some of the most important developments in computer technology (for example, what was once widely known as 'time sharing”) have been furthered by quantum chemists. This lecture will look at developments in both fields of endeavor from 1940 to the present.

Abstract: Beginning with the famous 1929 statement of Paul Dirac concerning the complexity of the application of the fundamental equations of quantum mechanics to chemistry, computers and quantum chemistry have been joined at the hip. Indeed some of the most important developments in computer technology (for example, what was once widely known as 'time sharing”) have been furthered by quantum chemists. This lecture will look at developments in both fields of endeavor from 1940 to the present.

Abstract: Metropolis Monte Carlo simulations of the eutectic NaK alloy are performed using the Second Moment Approximation (SMA) model potential across a wide range of temperatures at constant pressure. The alloy structure and thermodynamics are analyzed along with the atomic level structures using a variety of structure identification methods. Both enthalpy and density are followed along an annealing process that reveals a clear melting point around 260 K. At lower temperatures, two thermodynamic branches are identified as crystalline and amorphous solids.

Abstract: In this work we have developed a multi-tiered computational platform to study protein-drug interactions. At the beginning of the workflow more efficient and less accurate methods are used to enable large libraries of proteins in many conformations and massive chemical libraries to be screened. At each subsequent step in the workflow a subset of input data is investigated with increased accuracy and more computationally expensive methods. In this work we demonstrate the developed workflow with the investigation of the lymphocyte-specific kinase LCK, which is implicated as a drug target in many cancers and also known to have toxic effects when unintentionally targeted.

Abstract: Hypervalent iodine compounds, in particular trivalent iodine species (lambda-3 iodanes) have become important reagents for the transfer of electrophilic substituents to arenes and other nucleophiles. Quantum chemical studies have shown that iodanes have a rich and often complex reactivity, reminiscent of the one of organometallics. We further explored the reactivity of these compounds based on data-centric models. A large array of iodanes with different electrophilic substituents, different leaving groups and bridges was created and investigated statistically. In this presentation, we show that the kinetic and thermodynamic stability of these reagents can be described by simple structural parameters such as bond lengths. This allows us to develop a stability test for these reagents by using a machine learning technique, e.g., the support vector machine.

Abstract: Interest in simulation based screening of soft materials for desired chemical and physical properties has grown in recent years, which promises a reduced time to discovery and significantly lower costs than experimentation. While programmatically generating simulator inputs is fairly straightforward for crystals or small molecules, this is still challenging for complex (connected) molecular systems in general, and soft materials in particular. To tackle this challenge, this work presents a domain specific modeling environment that (1) features a domain specific modeling language (DSML), capable of describing classes of molecular systems in a hierarchical, component based model, with a rich set of composition operators, and support for parameterizable systems though generative modeling; (2) a browser-based, intuitive graphical user interface; (3) support for online collaboration, sharing of components, support for version control with history and branching; and (4) an interpreter to visualize components and molecular systems, and to generate output understood by major molecular dynamics (MD) simulator tools. The paper includes a case study demonstrating the use of the modeling environment to build a generative model of a parameterizable nanostructure, to feed input to a simulation based screening workflow of nanolubrication materials.

Abstract: Microscopic structure of a fully hydrated di-palmytoil-phosphatidyl-choline lipid bilayer mem- brane in the liquid-crystalline phase has been analyzed with all-atom molecular dynamics sim- ulations based on the recently parameterized CHARMM36 force field. Within the membrane a single molecule of the alpha-aminoacid tryptophan (precursor of important neurotransmitters such as serotonin and melatonin) has been embedded and his structure and binding sites have been explored. In addition, properties such as radial distribution functions, energy and pressure profiles and the potentials of mean force of water-tryptophan and lipid-tryptophan have been evaluated. It has been observed that tryptophan tends to be close to the lipid headgroups but that it can be fully hydrated during short time intervals of the order of 1 ns. This would indicate that hydrophobic forces as well as the attraction of tryptophan to polar sites of the lipids play a significant role in the definition of the structure and binding states of tryptophan.