TY - JOUR
T1 - A knowledge-based weighting framework to boost the power of genome-wide association studies
AU - Li, Miao Xin
AU - Sham, Pak C.
AU - Cherny, Stacey S.
AU - Song, You Qiang
PY - 2010
Y1 - 2010
N2 - Background: We are moving to second-wave analysis of genome-wide association studies (GWAS), characterized by comprehensive bioinformatical and statistical evaluation of genetic associations. Existing biological knowledge is very valuable for GWAS, which may help improve their detection power particularly for disease susceptibility loci of moderate effect size. However, a challenging question is how to utilize available resources that are very heterogeneous to quantitatively evaluate the statistic significances. Methodology/Principal Findings: We present a novel knowledge-based weighting framework to boost power of the GWAS and insightfully strengthen their explorative performance for follow-up replication and deep sequencing. Built upon diverse integrated biological knowledge, this framework directly models both the prior functional information and the association significances emerging from GWAS to optimally highlight single nucleotide polymorphisms (SNPs) for subsequent replication. In the theoretical calculation and computer simulation, it shows great potential to achieve extra over 15% power to identify an association signal of moderate strength or to use hundreds of whole-genome subjects fewer to approach similar power. In a case study on late-onset Alzheimer disease (LOAD) for a proof of principle, it highlighted some genes, which showed positive association with LOAD in previous independent studies, and two important LOAD related pathways. These genes and pathways could be originally ignored due to involved SNPs only having moderate association significance. Conclusions/Significance: With user-friendly implementation in an open-source Java package, this powerful framework will provide an important complementary solution to identify more true susceptibility loci with modest or even small effect size in current GWAS for complex diseases.
AB - Background: We are moving to second-wave analysis of genome-wide association studies (GWAS), characterized by comprehensive bioinformatical and statistical evaluation of genetic associations. Existing biological knowledge is very valuable for GWAS, which may help improve their detection power particularly for disease susceptibility loci of moderate effect size. However, a challenging question is how to utilize available resources that are very heterogeneous to quantitatively evaluate the statistic significances. Methodology/Principal Findings: We present a novel knowledge-based weighting framework to boost power of the GWAS and insightfully strengthen their explorative performance for follow-up replication and deep sequencing. Built upon diverse integrated biological knowledge, this framework directly models both the prior functional information and the association significances emerging from GWAS to optimally highlight single nucleotide polymorphisms (SNPs) for subsequent replication. In the theoretical calculation and computer simulation, it shows great potential to achieve extra over 15% power to identify an association signal of moderate strength or to use hundreds of whole-genome subjects fewer to approach similar power. In a case study on late-onset Alzheimer disease (LOAD) for a proof of principle, it highlighted some genes, which showed positive association with LOAD in previous independent studies, and two important LOAD related pathways. These genes and pathways could be originally ignored due to involved SNPs only having moderate association significance. Conclusions/Significance: With user-friendly implementation in an open-source Java package, this powerful framework will provide an important complementary solution to identify more true susceptibility loci with modest or even small effect size in current GWAS for complex diseases.
UR - http://www.scopus.com/inward/record.url?scp=79251484337&partnerID=8YFLogxK
U2 - 10.1371/journal.pone.0014480
DO - 10.1371/journal.pone.0014480
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
C2 - 21217833
AN - SCOPUS:79251484337
SN - 1932-6203
VL - 5
JO - PLoS ONE
JF - PLoS ONE
IS - 12
M1 - e14480
ER -