startEWAS.RdPerform EWAS analysis to obtain the coefficient value, standard deviation and significance p value (or adjust p value) of each site.
startEWAS(input, filename ="default", model = "lm", expo = "default",
cov = NULL,random = NULL, time = NULL, status = NULL, chipType = "EPICV2",
adjustP = TRUE, core = "default")An R6 class integrated with all the information obtained from the loadEWAS or transEWAS function.
User-customized .csv file name for storing EWAS results. If "default" is chosen, it will be named as "ewasresult".
The statistical models used for EWAS analysis include "lm" (general linear regression), "lmer" (linear mixed-effects model), and "cox" (Cox proportional hazards model). The default model is "lm".
Name of the exposure variable used in the EWAS analysis.
Name(s) of covariate(s) used in the EWAS analysis, with each name separated by a comma. Ensure there are no space. e.g. "cov1,cov2,cov3".
Random intercept item name, used only when selecting the "lmer" model.
When the user selects the Cox proportional risk model, the name of the time variable needs to be specified.
When the user selects the Cox proportional risk model, the name of the status variable needs to be specified.
Whether to calculate adjusted p-values(FDR and Bonferroni correction). The default is set to TRUE.
The Illumina chip versions for user measurement of methylation data, including "27K", "450K","EPICV1", "EPICV2", and "MSA". The default is "EPICV2".
The number of cores used during parallel computation. If set to default, it calculates the maximum number of available physical cores minus 1 and treats this as an operational kernel.
input, An R6 class object integrating all information.
if (FALSE) { # \dontrun{
res <- initEWAS(outpath = "default")
res <- loadEWAS(input = res, ExpoData = "default", MethyData = "default")
res <- transEWAS(input = res, Vars = "cov1", TypeTo = "factor")
res <- startEWAS(input = res, filename = "default", chipType = "EPICV2", model = "lm",
expo = "var", cov = "cov1,cov2",adjustP = TRUE, core = "default")
} # }