Description Usage Arguments Details Value Author(s) Examples
Tests a set of variants represented in the numberical matrix G, given disease status and previously estimated probabilities of disease.
1 | BRtest(d, probs, G)
|
d |
An n vector of disease status |
probs |
An n vector of (estimated) disease probabilities corresponding to the individuals with disease status. |
G |
An nxp matrix of counts of (assumed rare) p variants genotpyed on n individuals. Columns correspond to variants, rows to individuals. |
For each variant, calculates the expected number of disease individuals among the carriers, and calculates p-value from the BinomiRare test of association between the variant and the disease status.
The function returns a data frame with variant name, the number of carriers of the variant, number of diseased carriers, the expected number of diseased carriers (according to the supplied probabilities), and p-value.
Tamar Sofer
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 | require(poibin)
########## Example 1: a single data set.
########## Simulate data
n <- 10000
effect.size <- 1
pop.risk <- -2.6
x <- rnorm(n, sd = 0.01)
x <- pmax(x, 0)
g <- rbinom(n, size = 2, prob = x) ## one causal variant, x is a confounder
G <- matrix(rbinom(n*100, size = 1, prob = 0.001), nrow = n) ### another 100 null variants
G <- cbind(g, G)
colnames(G) <- paste0("var_", 1:ncol(G))
rownames(G) <- paste0("person_", 1:nrow(G))
p <- expit(pop.risk + g*effect.size + 20*x)
d <- rbinom(n, 1, p)
names(d) <- paste0("person_", 1:nrow(G))
########### Now that we have outcome d, genotypes G and a covariate x:
########### Estimate disease probability model
prob.mod <- glm(d ~ x, family = binomial)
prob.d <- expit(predict(prob.mod))
system.time(res <- BRtest(d, prob.d, G)) ### super quick
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.