Supplementary Material for: On Rare-Variant Analysis in Population-Based Designs: Decomposing the Likelihood to Two Informative Components
datasetposted on 14.01.2014 by Won S., Kim Y., Lange C.
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Various analytical approaches have been suggested for the characterization of rare variants. One main approach is to collapse the genetic information of rare variants in a region and to construct an overall test statistic. Here, we proposed a new approach based on collapsed genotype scores. By utilizing the information of the association signal that is ignored in collapsing methods, i.e. the configuration of rare alleles, we constructed a more powerful test and compared it with existing rare-variant approaches. With extensive simulation studies, we showed that our method performs better than existing approaches, and we applied our method to a sequencing study of nonsyndromic cleft lip illustrating the practical advantages of the proposed method.