The Metadata-Driven Online Registry Generator (mdORG) developed in collaboration with the University of Nebraska Medical Center Fred & Pamela Buffet Cancer Center, streamlines creation, maintenance and upgrades of disease-specific registries.
The mdORG consists of:
MS Excel template to maintain the registry’s metadata
Metadata Loading Engine (MLE) to build registry-specific metadata schema in the Oracle database
Automated Registry Builder (ARB) to deploy the registry’s web application and to dynamically build or alter the registry's database based on the defined metadata.
A Registry generated by the ARB is a multi-tier web application that contains:
back-end (Oracle database schema)
middle tier (Java servlets),
runtime data manipulation library
Utilizing our novel mdORG tool, researchers can build registries that are: (i) easily upgradeable and maintainable; (ii) compatible with all major browsers; (iii) adaptable to end-user devices (including Apple iPads and Android tablets); and (iv) compliant with modern standards for data collection and management.
The iCON is a web application that allows to securely obtain patient’s informed consent electronically and store it in a PDF format on a secure server. The iCON supports multiple versions of informed consents. It is compatable with all major web browsers and end-user devices, including Apple iPads and Android tablets.
PeptDB is a relational database on structural information on amino acid residues within proteins with known three-dimensional structures. This database is designed to provide bioinformatics support for solving scientific and applied problems requiring information on the spatial organization of proteins and peptides.
PeptDB contains information on the dihedral angles, a surface area accessible for water molecules (fractional area), energy of interaction of the atoms of the given residue with the atoms of other residues in the protein (energy by residue), number of other residues found in the 8 Å sphere, involvement of the residue in the formation of a particular type of secondary structure, adaptation of a particular type of conformational cluster, etc. The ultimate goal of the PeptDB is to continuously collect the corresponding information for all proteins with well-known 3D structures. The current release of the database contains records on 762,538 amino acid residues derived from 2,882 crystallographic protein structures with a sequence identity percentage cutoff of 30%, resolution better than 2.5 Å, and with an R factor less than 0.25.
(click the picture to enlarge)
PeptDB can be utilized for protein engineering, protein and peptide structure prediction, improving the accuracy and precision of protein and peptide structure determination from x-ray or nmr data, and serve the needs of homology modeling of proteins.
PeptDB can be useful for finding epitopes on the surface of proteins to be utilized in the rational design of imunogenes and antibodies.
FiSiNOE-3 aimed to determine the most probable local conformations of proteins and peptides based on the combined use of NMR data with information on local structures extracted from Protein Data Bank (PDB). The program uses, as a priori data, a new set of clusters for , , and angles, obtained from the joint probability density functions of proteins with a high resolution from PDB. By a given set of geometrical restrictions derived from NMR data, the program has estimated posterior probabilities, mathematical expectations, and standard deviations for , , and angles. To do this, the program uses the Bayesian classification method across the clusters. The clusters having posterior probabilities that exceed of a given significance level, FiSiNOE-3 also estimates combined ranges for , , and angles with a given half-width of the confidence interval. As input data, the program uses: a file with an amino acid sequence; a file with estimated values for upper limits of interproton distances (in angstroms), corresponding to intraresidue and sequential cross-peak intensities of nuclear Overhauser effects; a file with values of coupling constants 3JN and 3J ß (in Hz); a value of a half-width of the confidence interval; and a value of significance level for probabilities.
As output, the program creates the following files: a file with acceptable conformations for amino acid residues (posterior probabilities, mathematical expectations, and standard deviations for , , and angles); and a file with limits for torsion angles (allowed intervals for , , and angles, combined for all clusters, for which a posterior probability exceeds a given significance level). The formats of input and output files of FiSiNOE-3 are compatible with the corresponding files used in HABAS and DIANA programs, widely used for protein structure determination from NMR data. FiSiNOE-3 is written in C++ and compiled under IRIX and Solaris operating systems.
The ConCoder automatically encodes 3D structures of proteins into their 1D conformational profiles. The 1D profile is stated using a short alphabet to code the typical conformational states that may be adopted by amino acid residues in protein structures. The alphabet was obtained by mixture-model cluster analysis of torsion angle values derived from spatial structures of proteins. The set of typical conformational states (clusters) for amino acid residues was determined from a selected set of non-redundant high-resolution protein structures. Membership to a particular conformational cluster is determined by the following maximum a posteriori probability (MAP) procedure: Given the torsion angles of a protein (derived from 3D structure), the procedure estimates the probabilities for each amino acid residue within the given protein to belong to each allowed conformational cluster. The distributions of the torsion angles within a cluster are approximated by Gaussian distributions.
The log posterior probability of membership in a given cluster is determined essentially by the corresponding Mahalanobis D-square (squared statistical distance) between the observed values of the torsion angles of the residue and the mean of the distribution for a particular conformational cluster, in the metric of the covariance matrix of that distribution, plus the log of the determinant of that covariance matrix. The conformation of the residue is assigned to the conformational cluster with which there is maximum a posterior probability. As input data, the program uses the spatial structure of a protein, written in the PDB format, and encodes this 3D structure into 1D sequence of conformational states (clusters) adopted by amino acid residues in the protein.
PGX Webmaster™ is the web content management solution that enables authorized non-technical staff to make the changes necessary to keep company website up to date, saving company time and money.
PGX Webmaster provides an easy-to-use, highly affordable web maintenance tool that can save companies thousands of dollars in both internal and external technical resources. Content contributors are empowered to quickly make time-sensitive changes and to have control over the content that they are ultimately responsible for generating and maintaining.
Turbomarket Distribution Portals
The TurboMarket Distribution Portals allows a business of any size to have its own website with a centrally hosted distributor-branded storefront that allows clients to place orders electronically or by using the help of a operator. By automating routine tasks, distributors substantially reduce business expenses and have more time to build relationships with current and potential customers. The goal of this application is to provide the eServices that allow small to medium size distributors to interact effectively with the buyers and suppliers via the Internet.
Progenomix was involved in the development of this web-based application and currently represents TurboMarket Corporation in Nebraska.