LONDON — Gemini Research Ltd., a genomics company that specializes in discovering disease genes through the study of twins, has agreed to a research collaboration with Chiroscience Group plc to identify genes associated with obesity. The work will be carried out by Cambridge-based Chiroscience's genomics subsidiary, Rapigene Inc., of Seattle.

Under terms of the collaboration, Gemini, also of Cambridge, will select genes relevant to obesity and identify candidate genetic markers, or single-nucleotide polymorphisms (SNPs). Rapigene will then incorporate these candidate SNPs into its proprietary SNP detection assay. Gemini will determine whether the candidate SNPs are associated with obesity by correlating the genotype data with phenotype data from its twins database, called Phenobase.

Paul Kelly, CEO of Gemini, told BioWorld International the database is "a very powerful resource for validating genetic information. Combined with Rapigene's high-throughput screening system, [it's] a well-analyzed database of humans [joined] with a technology which offers the potential to identify important disease genes and drug targets in obesity."

Database Covers More Than 500 Variables

The two companies will bear their own costs. "Any intellectual property we generate will be jointly commercialized," said Kelly. "It could be that it is taken up by Chiroscience itself, or that we will find other partners." Any income generated will be shared.

Kelly stressed the deal is nonexclusive. Gemini already has an agreement with Affymetrix Inc., of Santa Clara, Calif., to apply its high-throughput screening technology to find genes associated with osteoporosis in the Gemini database. "Both of these partners have very powerful technologies. One of the things we are looking at [in the collaborations] is how best to extract data from our database," Kelly said. "We would like to look at all the available screening technologies."

While Gemini is funding its own portion of the work, Kelly said, there are deals in the pipeline in which the company will be paid for access to its database. Over the past six years, Gemini has built up an extensive clinical collection based on comparing the phenotypes of identical twins. The company's database now covers more than 500 clinical variables which, it says, allow rapid and efficient identification and validation of genes associated with known risk traits in disease. *