gxe {Haplin}

Test for gene-environment interaction

Description

Performes a gene-environment test to check if haplin estimates of relative risks change over strata of environment. It is typically applied to the output from haplinStrat

Usage

gxe(object.list)

Arguments

object.list

A list of haplin results, almost always the output from haplinStrat. The first element is the result of running haplin on all data; the remaining elements are the results for each stratum separately.

Details

haplinStrat runs haplin first on the entire input data file, then on each stratum separately. The results from haplinStrat are similar to just manually splitting the file into strata and running haplin on each, with one important difference, however: Since some strata may be small etc., haplin might conceivably choose different haplotypes in different strata, and also choose different reference haplotypes. When first running haplin on the entire file, haplinStrat saves the selected haplotypes and chosen reference category. Then, in the strata-specific runs haplinStrat forces haplin to choose the same haplotypes/reference category in all runs, so that results from different strata are comparable. When applying gxe to the output from haplinStrat, it will test whether there is a statistically significant change in parameter estimates from stratum to stratum, i.e. a gene-environment interaction since strata usually are defined by an environmental exposure. gxe uses Wald tests to test for interactions. It always tests whether there is change in haplotype frequencies from stratum to stratum. More importantly, it separately tests whether any genetic effects, such as fetal genetic effects, maternal effects, or parent-of-origin effects, change significantly over strata. gxe can also be run from within haplinSlide by using the strata argument in haplinSlide.

Value

A dataframe with one row for each test that is performed (haplo.freq is the first, the remaining depend on the model that has been estimated). The Wald chi-squared test value, degrees-of-freedom, and resulting p-value are reported.

NOTE:

In the future, the structure of the output from gxe will change. In particular, measures of ratios of relative risks will be reported in addition to the p-values

Note

Further information is found on the web page.

Author(s)

Hakon K. Gjessing
Professor of Biostatistics
Division of Epidemiology
Norwegian Institute of Public Health
[e-mail]

References

Gjessing HK and Lie RT. Case-parent triads: Estimating single- and double-dose effects of fetal and maternal disease gene haplotypes. Annals of Human Genetics (2006) 70, pp. 382-396.

Web Site: https://haplin.bitbucket.io

See Also

haplin, haplinStrat, haplinSlide

Examples

## Not run: 
# All standard haplin runs can be done with haplinStrat. 
# Below is an illustration. See the haplin help page for more 
# examples.
# 
# Analyzing the effect of fetal genes, including triads with missing data,
# using a multiplicative response model. The first column of the data file
# in this example contains the stratification variable.
result <- haplinStrat("C:/work/data.dat", strata = 1, use.missing = T, response = "mult",
reference = "ref.cat", winlength = 1)
# Provide summary of separate results:
lapply(result, summary)
# Plot results separately:
par(ask = T)
lapply(result, plot)
#
# Convert results to table format and stack them over strata:
haptable(result)
# Test for interaction between haplotype risk estimates and the strata variable:
gxe(result)


## End(Not run)