Molecular Computational Core Facility
Many of the MCCF promoted investigations are associated with two UM Centers, including Structural and Functional Neuroscience (CSFN) and Biomolecular Structure and Dynamics (CBSD). The computational facility supports both long- and short-term investigations. It promotes the use of open-source software in companion to proprietary software packages employing Linux-Unix-OSX seamlessly integrated platforms.
The MCCF has facilitates homology-based biopolymer (protein, DNA, RNA) structure and function modeling, electronic structure calculations, classical molecular dynamics (MD) and Monte Carlo sampling, stochastic dynamics, molecular docking, and electrostatic-steric analyses. Combinations of these studies provide the best possible inference of structure-function relationships for biomolecules, ligands and their interactive dynamic forms. Additionally, homology and motif search tools have been integrated with phylogenic profiling to advance evolutionary-based protein models thereby advancing subsequent ab initio molecular dynamic simulations and de novo folding methods for loop modeling and related studies.
The investigations have established benchmark standards for quality assessments, and uncertainty estimations of the models, maps and MD simulated trajectory solutions. Parallel processing, code development, and the fabrication with use of relational databases (composed of molecular-biopolymer descriptors in relation to laboratory biological-biochemical-spectroscopic data sets) come forward to various extents.
Other focus areas encompass the fabrication of computationally derived models; viz., multiple-attribute correlated three-dimensional (3D) predictive protein-ligand bound maps. These contemporary maps are composed where the attributes are critical measures of molecular biological protein compositions, interacting ligand constitutions, spectroscopic, kinetic and pharmacological qualities of protein-ligand relationships and related biophysical data.
For the most part, the requisite model correlated data sets are acquired from other UM core facilities (X-ray crystallography, Biospectroscopy, Mass Spectral, P30-Core Lab for Neuromolecular Production, Histology, amongst others) and individual investigator laboratories (e.g., pharmacological potency-selectivity measures, NMR, fluorescent spectroscopic characterizations, patch clamp assessments, kinetic profiling, amongst others).
Together, the multidiscipline derived structure-function protein and ligand maps, models and simulations provide keen insights into protein structure-function and routinely drive the generation of new molecular and biopolymer hypotheses.
The Extreme Science and Engineering Digital Environment (XSEDE; formerly TeraGrid) offers free high-performance and high-throughput computational resources at NSF-sponsered supercomputer sites. Full researcher allocations are granted through a peer-reviewed merit-based application process (http://www.xsede.org/). The MCCF promotes efforts to utilize the advanced computing power of the NSF sponsored XSEDE. Prospective XSEDE jobs are first evaluated locally for initial algorithm code performance, parameter optimizations and data stream output storage handling using either the UM Altix node cluster operated with the aid of Mr. J. Glassy, System Administrator or high powered workstation machines. Subsequently, with a refined protocol the investigator submits jobs to XSEDE machines with assistance of the MCCF application specialist and XSEDE user, Dr. M. Braden.