US Patent 7,249,116. Machine learning. Inventor: Michael Stephen Fiske. Assignee: Fiske Software.
Protects methods for representing patterns, information, and knowledge in computational applications. According to the patent abstract, the invention has applications in bioinformatics, document classification, document similarity, financial data mining, goal-based planners, handwriting and character recognition, information retrieval, natural language processing and understanding, pattern recognition, search engines, strategy based domains such as business, military and games, and vision recognition.
US Patent 7,248,971. Method and apparatus for discovering patterns in a set of sequences. Inventors: Isidore Rigoutsos, Yuan Gao, Aristidis Floratos. Assignee: International Business Machines.
Covers an unsupervised approach for identifying additional members of a family that is defined initially through exemplar sequences. The invention “proceeds without any information related to the exemplar sequences defining the family, without aligning the sequences, without prior knowledge of any patterns in the exemplar sequences, and without knowledge of the cardinality or characteristics of any features that may be present in the exemplar sequences,” according to the patent abstract. In one aspect of the invention, a method takes a set of unaligned sequences and discovers several or many patterns common to some or all of the sequences. These patterns can then be used to determine if candidate sequences are members of the family.
US Patent 7,243,112. Multidimensional biodata integration and relationship inference. Inventors: Kunbin Qu, Nan Lin, Yanmei Lu, Donald Payan. Assignee: Rigel Pharmaceuticals.
Protects a platform for analyzing biological data that “emphasizes pathway mapping and relationship inference based upon data acquired from multiple diverse sources.” The platform is based on a bioinformatics system that integrates data from diverse sources, connects related genes and proteins, and infers biological functions in the context of cellular processes.
US Patent 7,243,051. System for predicting three-dimensional structure of protein. Inventor: Kentaro Tomii. Assignee: National Institute of Advanced Industrial Science and Technology.
Protects a system for measuring the similarity between protein profile matrices that is useful for predicting a protein three-dimensional structure. The invention provides a system for measuring the similarity between protein profile matrices, in which the matrix includes a group of profile columns containing occurrence probabilities for every amino acid type in a multiple alignment. The system for measuring the similarity includes a means for preparing two profile matrices of a query profile matrix and a subject profile matrix; a means for calculating correlation coefficients between the profile columns in the query and subject profile matrices with respect to full or partial combinations of both the respective profile columns; and a means for forming a score matrix consisting of the correlation coefficients.
US Patent 7,243,031. Method of designing multifunctional base sequence. Inventors: Yoko Satou, Masato Kitajima., Kiyotaka Shiba. Assignees: Fujitsu, Kiyotaka Shiba.
Protects a computational method for designing a multifunctional base sequence in which a base sequence has two or more functions in different reading frames, the sequence data of a protein or a peptide encoded by a base sequence arising from one of the three reading frames is processed as a pool of oligopeptide units, and the base sequence information arising from other reading frames contained in the oligopeptide sequence is used.
US Patent 7,240,042. System and method for biological data analysis using a Bayesian network combined with a support vector machine. Inventors: Jie Cheng, Chao Yuan, Bernd Wachmann, Claus Neubauer. Assignee: Siemens Medical Solutions USA.
Protects a method for analyzing biological data that includes classifying a set of biological data with one classifier, classifying a second set of biological data in a second classifier, combining the results of the first classifier with the results of the second classifier, and analyzing the results as a function of the similarity measure of the first classifier and the similarity measure of the second classifier.
US Patent 7,240,038. Heuristic method of classification. Inventors: Ben Hitt. Assignee: Correlogic Systems.
Protects heuristic algorithms for classifying objects. A genetic algorithm is used to abstract a data stream associated with each object and a pattern recognition algorithm is used to classify the objects and measure the fitness of the chromosomes of the genetic algorithm. The learning algorithm is applied to a training data set. The learning algorithm generates a classifying algorithm, which is used to classify or categorize unknown objects. According to the patent abstract, the invention is useful in the areas of classifying texts and medical samples, predicting the behavior of financial markets, and in monitoring the state of complex process facilities to detect impending failures.