Looking for the rare gene variants that may be associated with human disease can be complicated by unknown biases in the makeup of populations, resulting in spurious associations.

A group of researchers at Baylor College of Medicine (www.bcm.edu) and the University of Washington in Seattle have found that using so-called trio analysis, analyzing the genomes people with disease and those of their parents using the transmission disequilibrium test can reduce these problems. (The transmission disequilibrium test detects associations (when the two occurrences of traits, of which at least one is genetic, in a population occurs too frequently to be attributed to chance)  only in the presence of genetic linkage (the tendency of genes located near each other on a chromosome to be inherited together during meiosis or the sexual division of cells) only in the presence of genetic association. They then used different statistical methods to further modify this test to analyze rare variants obtained from next generation sequence data.  A report of their work appears online in the American Journal of Human Genetics.

“Sometimes you find false associations when you are analyzing these data because of the population substructure,” said Dr. Suzanne M. Leal, professor of molecular and human genetics and director of the Center for Statistical Genetics at BCM. 

“Your cases and controls are coming from different populations and you do not realize it. Or you can have different disease frequencies in different populations.” However using the rare variant transmission disequilibrium test, these false associations are avoided. (Leal is corresponding author of the report.)

For example, studies in Pima Indians, known for high rates of type 2 diabetes and hypertension, can be misleading because they have married with people of European descent. If you compare a population of such Indians that have type 2 diabetes with those that do not, it is likely that those who are closer to being pure Pima will be in the study group and the control group will have more people of European descent. 

In their study, Leal and her colleagues analyzed exome sequence data in the Simon Simplex Collection. The study involved 199 trios that included one person who had autism and his or her parents. Their study identified variants in a gene called ABCA7 that might be associated with autism – one of many such so-called autism genes. Common variants in this gene have been associated with Alzheimer’s disease in previous studies.

“The identification of ABCA7 as a gene that is possibly involved in autism etiology suggests the existence of a common pathway for neurodevelopmental and neurodegenerative diseases that may be targeted for prevention and treatment,” said the researchers.

Others who took part in this work include: Zongxiao He, Gao Wang, Stanley Hooker, Regie Lyn P. Santos-Cortez, Biao Li and Mengyuan Kan, all of BCM; Brian J. O’Roak, Joshua D. Smith, Nik Krumm, Deborah A. Nickerson, Jay Shendure and Evan E. Eichler, all of the University of Washington School of Medicine in Seattle.

Funding for this work came from the Simons Foundation Autism Research Initiative (Grants SFARI 137578 and 191889), the Howard Hughes Medical Institute and the National Institutes of Health (Grants HD065285, HL102926, MD005964 and HG006493).