TY - GEN
T1 - A biologically informed method for detecting associations with rare variants
AU - Buchanan, Carrie C.
AU - Wallace, John R.
AU - Frase, Alex T.
AU - Torstenson, Eric S.
AU - Pendergrass, Sarah A.
AU - Ritchie, Marylyn D.
PY - 2012
Y1 - 2012
N2 - With the recent flood of genome sequence data, there has been increasing interest in rare variants and methods to detect their association to disease. Many of these methods are collapsing strategies which bin rare variants based on allele frequency and functional predictions; but at this point, most have been limited to candidate gene studies with a small number of candidate genes. We propose a novel method to collapse rare variants based on incorporating biological information from the public domain. This paper introduces the functionality of BioBin, a biologically informed method to collapse rare variants and detect associations with a particular phenotype. We tested BioBin using low coverage data from the 1000 Genomes Project and discovered appropriate binning characteristics based on what one might expect given the size of the gene. We also tested BioBin using the pilot targeted exome data from 1000 Genomes Project. We used biologically-informed binning and differences in minor allele frequencies as a means to distinguish between two ancestral populations. Although BioBin is still in developmental stages, it will be a useful tool in analyzing sequence data and uncovering novel associations with complex disease.
AB - With the recent flood of genome sequence data, there has been increasing interest in rare variants and methods to detect their association to disease. Many of these methods are collapsing strategies which bin rare variants based on allele frequency and functional predictions; but at this point, most have been limited to candidate gene studies with a small number of candidate genes. We propose a novel method to collapse rare variants based on incorporating biological information from the public domain. This paper introduces the functionality of BioBin, a biologically informed method to collapse rare variants and detect associations with a particular phenotype. We tested BioBin using low coverage data from the 1000 Genomes Project and discovered appropriate binning characteristics based on what one might expect given the size of the gene. We also tested BioBin using the pilot targeted exome data from 1000 Genomes Project. We used biologically-informed binning and differences in minor allele frequencies as a means to distinguish between two ancestral populations. Although BioBin is still in developmental stages, it will be a useful tool in analyzing sequence data and uncovering novel associations with complex disease.
UR - http://www.scopus.com/inward/record.url?scp=84859145177&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84859145177&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-29066-4_18
DO - 10.1007/978-3-642-29066-4_18
M3 - Conference contribution
AN - SCOPUS:84859145177
SN - 9783642290657
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 201
EP - 210
BT - Evolutionary Computation, Machine Learning and Data Mining in Bioinformatics - 10th European Conference, EvoBIO 2012, Proceedings
T2 - 10th European Conference on Evolutionary Computation, Machine Learning and Data Mining in Bioinformatics, EvoBIO 2012
Y2 - 11 April 2012 through 13 April 2012
ER -