Recent Patents in Bioinformatics, August — September 2011
US Patent 8,019,553. Method of modeling for drug design, evaluation, and prescription in the treatment of disease. Inventor: Michael Goldstein.
Describes an immunodynamic model that simulates the interactions between infection, immune response, and treatment in infectious diseases.
US Patent 8,014,953. RNA surveillance among curated proteins. Inventors: Steven E. Brenner, Richard E. Green, Tyler R. Hillman. Assignee: The Regents of the University of California.
The patent describes computational methods for characterizing putative protein isoforms as targets of nonsense-mediated decay. The program is used to identify mRNA sequences that represent transcripts that encode a set of protein isoforms; determine the corresponding gene intron-exon structures by mapping the mRNA sequences to their corresponding genomic sequences; and then determine if the transcripts are targets of NMD.
US Patent 8,005,627. Bioinformatic approach to disease diagnosis. Inventor: Richard Porwancher.
Describes a method of constructing a multivariate predictive model to diagnose disease using results from multiple diagnostic tests. The system involves selecting tests to include in a diagnostic panel. It uses an algorithm to weight the results of each test and generate a likelihood ratio, and then uses a clinical algorithm to estimate the pre-test probability of the disease based on clinical signs and symptoms. Finally, it combines both values to generate a post-test probability of the disease, which is compared to a cutoff value.
US Patent 8,005,626. System and computer readable medium for discovering gene regulatory models and genetic networks using relational fuzzy models. Inventors: Bhooshan P. Kelkar, Rajiv D. Bendale. Assignee: IBM.
Describes a system for building relational fuzzy models from sets of genes. The approach includes a data-selection method that clusters gene expression data and identifies a representative subset of genes and then uses a relational fuzzy modeling system to build a model of the gene set.