globalGSA (version 1.0)

globalSimes: Global Simes' combination method.

Description

This function provides the p-value for a joint test of association between a phenotype and a set of genetic variants (SNPs) using the Simes method [1] after a global test for the best mode of inheritance of every SNP. The final gene-p-value is obtained from the permutational null distribution of the test statistic

Usage

globalSimes(data, B, gene_list, Gene = "all", addit = FALSE,
covariable = NULL, family = binomial)

Arguments

data

Data frame containing the variables in the model. The first column is the dependent variable which must be a binary variable defined as factor (in case-control studies, the usual codification is 1 for cases and 0 for controls). SNP values may be codified in a numerical form (0,1,2) denoting the number of minor alleles, or using a character form where the two alleles are specified, without spaces, tabs or any other symbol between the two alleles.

B

Number of permutations considered in the permutational procedure.

gene_list

File that provides the name of the set (for instance, gene) where each SNP belongs. This file has two columns: the SNP-Id ("Id"), and the Gene-Id ("Gene"). The SNP-Id must have the same label as the colnames of the data file.

Gene

Name of the gene that we want to analyze. The default value is Gene= "all" that indicates that the p-values of all SNPs in the database are to be combined. In this case it is not necessary to specify the gene_list file. In other case, we need to specify the name of the gene, for instance, Gene = "Gene1", and also the gene_list file.

addit

logical to determine if only an additive inheritance model should be considered in the global Test or, conversely, if we want to consider all possible inheritance models (dominant, recessive, log-additive and co-dominant). By default, addit = FALSE.

covariable

Data frame containing the covariables in the model. Each column represents one covariable. By default, covariable=NULL.

family

This can be a character string naming a family distribution. By default, family=binomial.

Value

List with the following components:

nPerm

Number of permutations.

Gene

Considered Gene.

genevalue

gene-pvalue.

References

[1] Simes, R.J. (1986). An Improved Bonferroni Procedure for Multiple Tests of Significance. Biometrika, 73, 751-754.

Examples

# load the included example dataset. 
# This is a simulated case/control study data set 
# with 2000 patients (1000 cases / 1000 controls) 
# and 10 SNPs, where all of them have 
# a direct association with the outcome:
data(data)
#globalSimes(data, B=1000,  Gene="all", addit=FALSE)

# it may take some time,
# hence the result of this example is included:
data(ans31)

# You can test:
globalSimes(data, B=1,  Gene="all", addit=FALSE)

# We consider that the first four SNPs 
# are included in "Gene1", 
# and the other six SNPs
# are included in "Gene2":
data(gene_list)
#globalSimes(data, B=1000, gene_list=gene_list, Gene="Gene1", addit=FALSE)

# it may take some time,
# hence the result of this example is included:
data(ans3)

# You can test:
globalSimes(data, B=1, gene_list=gene_list, Gene="Gene1", addit=FALSE)