NEW YORK, July 5 – In a typical sequence homology search, researchers looking to determine the function of a particular human gene try to match its sequence with a similar sequence in another, simpler organism’s genome, such as that of the mouse. Because many mouse genes are better understood, this provides clues as to what the human gene does. In many cases, however, when there is no direct relative in the mouse or other genome, this method breaks down.
To tackle this problem, Charles DeLisi and his colleagues in the Bioinformatics Graduate Program and Department of Biomedical Engineering at Boston University show in this week’s Proceedings of the Natural Academy of Sciences that another, more round-about method for linking genes can also provide strong clues as to the unknown gene’s function. The method, called fusion linking, adds to the battery of weapons genomics researchers can use to decipher what role a gene plays in an organism, scientists said.
In fusion linking, researchers search one organism’s genome for two distinct genes whose sequences—in another organism—are fused together as one gene. Researchers in previous studies had suggested that such a relationship between genes indicated that the genes are related in function, and DeLisi’s team set out to prove it.
To do this, the researchers assembled 30 microbial genomes, and identified every possible pair of genes that could be fused together as one gene in another organism. DeLisi and his colleagues also identified 15 broad categories of gene function that they could use to classify each of the gene pairs they studied.
After searching through a comprehensive sequence database, the researchers discovered that 72 percent of the gene pairs that existed as a fused gene in another organism had a similar function, such as transcription, signal transduction, and energy production. “Our results are basically a way to provide hypotheses for what genes do,” said Adnan Derti, one of the authors of the paper.
David Eisenberg, a scientist at the Molecular Biology Institute at the University of California, Los Angeles, who has also studied fusion linking, said that the results are predictable, but represent the most comprehensive test of the method to date. “I think [gene fusion] will be an important part of the arsenal of determining gene function and protein function,” he said.
In fact, Eisenberg added that the technology has already attracted interest from the private sector. UCLA has applied for a patent covering certain aspects of the technology, he said, and has exclusively licensed their intellectual property to Protein Pathways, of Los Angeles, a company developing computational methods to identify protein interactions.