| testGeneFamilies {ShotgunFunctionalizeR} | R Documentation |
Main test function
testGeneFamilies(data, method="Binomial", samples=c(1,2),
multtest="BH", ord=TRUE,
design.matrix=NULL, coef=1,
covariates=NULL, log.covariates=FALSE, groups=NULL,
overdisp=FALSE)
data |
A ShotgunFunctionalizeR data object (e.g. from
readGeneFamilies) |
method |
What statistical methods should be applied to the data? |
samples |
Which samples should be compared? Only applicable for procedures comparing a pair of metagenomes (currently the binomial, hypergeometric and Gaussian procedure) |
multtest |
What procedure should be used for correction for multiple
testing. See mt.rawp2adjp in the multtest package for
more information |
ord |
Should the resulting ranking list be sorted? |
design.matrix |
The design matrix. Only applicable when the general Poisson model is used |
coef |
Which coefficient from the design matrix should be tested? Only applicable when the general Poisson model is used |
covariates |
Covariates for poisson regression. Only applicable for regression analysis using the Poisson model |
log.covariates |
Should the covariates be logged? Only applicable for regression analysis using the Poisson model |
groups |
group labels for direct comparison. Only applicable for regression analysis using the Poisson model |
overdisp |
Should the quasi-Poisson family with overdisperson be used for fitting the GLM? Only applicable when the Poisson model is used |
readGeneFamilies,
testGeneFamilies.binomial,
testGeneFamilies.hypergeometric,
testGeneFamilies.gaussian,
testGeneFamilies.poisson,
glm,
family