A comprehensive evaluation of collapsing methods using simulated and real data: Excellent annotation of functionality and largesample sizes required

• Main Authors: Carmen eDering, Inke R König, Laura B Ramsey, Mary V Relling, Wenjian eYang, Andreas eZiegler
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2014-09-01
• Series: Frontiers in Genetics
• Subjects: comparison; rare variants; SLCO1B1; simulation study; collapsing; burden test
• Online Access: http://journal.frontiersin.org/Journal/10.3389/fgene.2014.00323/full

The advent of next generation sequencing (NGS) technologies enabled the investigation ofthe rare variant-common disease hypothesis in unrelated individuals, even on the genome-widelevel. Analysis of this hypothesis requires tailored statistical methods as single marker tests failon rare variants. An entire class of statistical methods collapses rare variants from a genomicregion of interest (ROI), thereby aggregating rare variants. In an extensive simulation study usingdata from the Genetic Analysis Workshop 17 we compared the performance of 15 collapsingmethods by means of a variety of pre-defined ROIs regarding minor allele frequency thresholdsand functionality. Findings of the simulation study were additionally confirmed by a real dataset investigating the association between methotrexate clearance and the SLCO1B1 gene inpatients with acute lymphoblastic leukemia. Our analyses showed substantially inflated type Ierror levels for many of the proposed collapsing methods. Only four of 15 approaches yieldedvalid type I errors in all considered scenarios. None of the statistical tests was able to detect trueassociations over a substantial proportion of replicates in the simulated data. Detailed annotationof functionality of variants is crucial to detect true associations. These findings were confirmedin the analysis of the real data. Recent theoretical work showed that large power is achievedin gene-bases analyses only if large sample sizes are available and a substantial proportion ofcausing rare variants is present in the gene-based analysis. Many of the investigated statisticalapproaches use permutation requiring high computational cost. There is a clear need for valid,powerful and fast to calculate test statistics for studies investigating rare variants.