I am analysing rare variants and I performed rare variant association testing using gene level tests (exmpl. SKAT).
I want to demontsrate that SNP based association tests are less powerfull that gene-based when rare variants are analysed. Which SNP based association test do you suggest?
My sample size is around 1 500, consisting of almost the same number of cases and controls.
•
written
by
Gene •
1 answer
At the most simple level, it is just a Chi square or Fisher test on a 2 x 2 contingency table of minor and major alleles: A: SNP dataset and Z Score
Linear regression (continuous traits) and logistic regression (categorical traits) are also used, and, with these, one can technically control for the effect of covariates, which may include other SNPs.
One can improve these and elaborate them to a genetic context by assuming a dominantly or recessively inherited disease allele - PLINK goes over these and even more tests: https://www.cog-genomics.org/plink/1.9/assoc
For rare variants, there are already 'burden tests', of which I am sure you are aware.
Kevin
Log in to answer this question.
-
SKAT vs SIMES methods
written by Eliza •
-
Simulation study to compare the power of different approaches
written by Gene •
-
Gene-based common and rare variant association analysis
written by seta
-
Extracting Cumulative (Rare) Allele Frequencies for a Subset of Variants From a VCF
written by maegsul
-
Single rare variant approaches
written by Tash •
-
PLINK/SEQ SKAT gives different p-values depending on number of tests run
written by p0078005 •
-
Whole exome sequencing data, rare variants and QQ-plots
written by felejohs
-
Rare variant association analysis for exome study without controls
written by mkaushal
-
Variant prevalence not listed
written by chloe.p.oconnell •
-
small case-control exome sequencing study
written by User6891
