US Patent 7,328,111. Method for determining similarities between data sequences using cross-correlation matrices and deformation functions. Inventor: Fatih Porikli. Assignee: Mitsubishi Electric Research Laboratories.
Protects a method for determining similarity between data sequences by comparing a first data sequence with a second data sequence and constructing a cross-correlation matrix of distances between all possible pairs of values of those sequences. The method then determines a minimum cost path between diagonally opposing corners of the cross-correlation matrix, and projects this path onto a transfer function that maps between the first data sequence and the second data sequence. The minimum cost path is then projected onto a deformation function that corresponds to a similarity between the first and second sequences. According to the patent claims, the method can be used in applications such as image retrieval, DNA pattern recognition, economic growth patterns, price and stock trends, weather and astrophysics patterns, and geological feature identification.
US Patent 7,324,926. Methods for predicting chemosensitivity or chemoresistance. Inventors: Pablo Tamayo, Jane Staunton, Donna Slonim, Hilary Coller, Todd Golub, Eric Lander, Jill Mesirov. Assignee: Whitehead Institute for Biomedical Research.
Protects methods for classifying or predicting the classes for samples based on gene expression, as well as methods for ascertaining or discovering new, previously unknown classes based on gene expression. The patent also describes methods and computer systems for predicting whether a sample is treatment sensitive or treatment resistant based on the analysis of gene expression data from samples that have been subjected to one or more compounds.
US Patent 7,321,830. Identifying drugs for and diagnosis of benign prostatic hyperplasia using gene expression profiles. Inventors: William Munger, Prakash Kulkarni, Robert Getzenberg, Iwao Waga, Jun Yamamoto. Assignee: Gene Logic.
Protects a computer-based method that includes a database “for elucidating changes in gene expression in prostate tissue isolated from patients exhibiting different clinical states of prostate hyperplasia as compared to normal prostate tissue,” according to the patent abstract. The method also identifies individual genes that are differentially expressed in diseased prostate tissue, the abstract states.
US Patent 7,319,945. Automated methods for simulating a biological network. Inventors: Bruce Shapiro, Eric Mjolsness, Andre Levchenko. Assignee: California Institute of Technology.
Covers methods and computer systems and programs for simulating a biological network. According to the patent abstract, the described methods “facilitate biological network simulations via automated equation generation based on the concept of a hierarchy of canonical forms that describe biological processes at various levels of detail.” There are two classes of canonical forms at each level of the hierarchy: the input canonical form, which is used to supply information to the program; and the output canonical form, which is produced by a simulator.