US Patent 7,007,001. Maximizing mutual information between observations and hidden states to minimize classification errors. Inventors: Nuria Oliver, Ashutosh Garg. Assignee: Microsoft.
Covers a "system and methodology to facilitate machine learning and predictive capabilities in a processing environment," according to the patent abstract. The patent claims a computer-implemented learning system that comprises a prediction component to determine one or more states based in part upon previous training data and sampled data; and a classification model that cooperates with the prediction component to determine the one or more states. Claims include the system, "further comprising an application of bioinformatics," which includes "comprising a task to at least one of annotate a sequence into exons and introns, and compare the results with a ground truth."
US Patent 7,006,925. Method of determining a base sequence for nucleic acid. Inventors: Makoto Hazama. Assignee: Shimadzu.
Protects a method of sequence determination for nucleic acid. The method removes noise before determining the base sequence from data obtained by electrophoresis using waveform shaping by Fourier transformation.
US Patent 7,006,924. Method and system for searching for relationships between base sequences in genes. Inventors: Katsuji Watanabe, Mitsuru Okuda. Assignee: Incorporated Administrative Agency National Agriculture and Bio-oriented Research Organization (Japan).
Protects a method and a system for searching for relationships between base sequences in order to efficiently analyze the relationship between a sample base sequence and known base sequences. In the system, "a theoretical value calculating portion" calculates theoretical restriction fragment length values based on restriction enzyme data, according to the patent abstract. A comparing portion then compares theoretical restriction fragment patterns with measured restriction fragment patterns, calculating the degree of similarity of the sample base sequence to known base sequences.
US Patent 7,006,921. Method for accurately estimating pKa of molecules using atom type definitions and partial least squares. Inventors: Li Xing, Robert Glen. Assignee: None.
Protects a computer-implemented software method for predicting the pKa of an arbitrary molecule based upon a knowledge of the molecular structure of that molecule and a statistical analysis of the molecular structures of a group of molecules for which the pKa is known.
US Patent 7,003,403. Quantifying gene relatedness via nonlinear prediction of gene. Inventors: Edward Dougherty, Seungchan Kim, Michael Bittner, Yidong Chen. Assignee: The United States of America as represented by the Department of Health and Human Services; The Texas A & M University System.
Protects a method for quantifying relatedness between genes by constructing nonlinear models predicting gene expression. Each gene's contribution to the measurement of relatedness can be shown on a graph, and graphical representations of models used to predict gene expression can be displayed.
US Patent 6,999,607. Method and apparatus for automated cellular bioinformatics. Inventors: Eskinder Kiros, Ping Ling, Robert Muller, Ramesh Venkat, Wesley Zink. Assignee: Cytokinetics.
Protects a system that conducts, monitors, and validates a cell-line-based biological experiment. The system comprises two principal subsystems, one of which is an automated data analysis manager for analyzing images, for analyzing the resulting image data to produce biological data, and for analyzing the biological data to produce biological phenotypes and treatment compound signatures.
US Patent 6,996,476. Methods and systems for gene expression array analysis. Inventor: Kayvan Najarian. Assignee: University of North Carolina at Charlotte
Covers methods and systems for applying independent component analysis and other signal-processing techniques to automatically identify an optimal number of independent gene clusters and to efficiently separate gene expression data into biologically relevant groups. The invention also includes an interface that allows the user to review the results at various stages during the analysis, and methods to mathematically define the relationship for gene expression within a group of interrelated genes.
US Patent 6,996,475. Computer software products for nucleic acid hybridization analysis. Inventor: David Balaban. Assignee: Affymetrix.
Protects computer software products for analyzing gene expression data. In one embodiment, multiple probes are used to detect a single transcript. The hybridization intensities of each probe are adjusted by dividing the intensities by the affinities of the probes. The minimal adjusted hybridization intensity may be used as the measurement of the gene expression.
US Patent 6,996,474. Automated method for identifying related biomolecular sequences. Inventors: Jacques Colinge, Rob Hooft Van Huijsduijnen. Assignee: Applied Research Systems ARS Holding.
Covers an automated method for identifying related biomolecular sequences having defined features of interest from databases. The methods includes the steps of: establishing from a first set of sequences a non-redundant list of query sequences having the defined features of interest; performing sequence alignments with the first family members in a second set of sequences derived from a second type of organism; establishing a two-dimensional matrix displaying the first and second family members and their respective similarity values; and selecting from the matrix those pairs of first and second family members for which the similarity values are the best among all of the alignments that involve one of the two pair's members.
US Patent 6,983,227. Virtual models of complex systems. Inventor: Cristina Thalhammer-Reyero. Assignee: Intertech Ventures.
Protects computer-based virtual models of complex systems, together with integrated systems and methods that provide a development and execution framework for visual modeling and dynamic simulation of said models. The virtual models in the present implementation "relate to biological complex systems," according to the patent abstract.