Abstract: The disclosure relates to novel approaches to mapping the MHC region and provides novel methods of genotyping the HLA loci. A haplotype map of the region and methods of using the map are also disclosed.

Abstract: The invention provides nucleic acid segments of the human genome, particularly nucleic acid segments from a gene, including polymorphic sites. Allele-specific primers and probes hybridizing to regions flanking or containing these sites are also provided. The nucleic acids, primers and probes are used in applications such as phenotype correlations, forensics, paternity testing, medicine and genetic analysis. A role for the thrombospondin gene(s) in vascular disease is also disclosed. Use of single nucleotide polymorphisms in the thrombospondin gene(s) for diagnosis, prediction of clinical course and treatment response, development of therapeutics and development of cell-culture-based and animal models for research and treatment are disclosed.

Abstract: The present invention features method for identifying an intestinal polyp according to the gen expression profile obtained from intestinal tissue. The present invention also features methods for identifying a compound that modulates intestinal polyp development, as well as oligonucleotide microarrays having genes involved in intestinal polyp development immobilized thereon.

Abstract: An array of oligonucleotides on a solid substrate is disclosed, which can be used for multiple purposes. Methods and reagents are provided for performing genotyping to determine the identity or ratio of allelic forms of a gene in a sample. A single base extension primer is coupled to a sequence identity code. During the primer extension reaction a distinctive label is incorporated which identifies the allelic form present in the sample. This permits multiple simultaneous analyses to be performed easily and efficiently.

Abstract: A neural network system and method for analyzing data sets, especially microarray gene expression data. The neural network is trained to generate time-dependent outcome predictions based on input features and outcome functions for a number of subjects. The features may be highly dimensional relative to the number of subjects, and feature selection is applied to the input feature data for training the neural network. A trained neural network processes input features from a subject to generate an outcome function that reflects the probability of the occurrence of an event at a given time point for that subject.

Abstract: Disclosed herein are methods and articles of manufacture for assembly of a plurality of reads from a genomic region that include the steps of indexing a plurality of read subsequences for each of a plurality of reads, extracting from the indexed subsequences a plurality of read pairs that have a predetermined number of subsequences in common, and merging the read pairs along a continuum. Also disclosed are methods and articles of manufacture for assembly of merged reads from a genomic region that include the steps of using linking information to identify one or more ambiguities in the merged reads, identifying a repeat region and a set of unique regions, and linking pairs of unique regions using the linking information associated with the reads in the unique regions.