curveFitting: Fit longitudinal data

In aametwally/OmicsLonDA: Omics Longitudinal Differential Analysis

Description

Fits longitudinal samples from the same group using negative binomial smoothing splines or LOWESS

Usage

curveFitting(formula = Count ~ Time, df = "NULL",
  fit.method = "ssgaussian", points = NULL)

Arguments

formula	formula to be passed to the regression model
df	dataframe has the Count, Group, Subject, Time
fit.method	fitting method (ssgaussian)
points	points at which the prediction should happen

Value

a list that contains fitted smoothing spline for each group along with 95

References

Ahmed Metwally ([e-mail])

Examples

library(SummarizedExperiment)
data("omicslonda_data_example")
omicslonda_se_object_adjusted = adjustBaseline(
                 se_object = omicslonda_data_example$omicslonda_se_object)
se_object = omicslonda_se_object_adjusted[1,]
dt = data.frame(colData(se_object))
dt$Count = as.vector(assay(se_object))
Group = as.character(dt$Group)
group.levels = sort(unique(Group))
gr.1 = as.character(group.levels[1])
gr.2 = as.character(group.levels[2])
df = dt
levels(df$Group) = c(levels(df$Group), "0", "1")
df$Group[which(df$Group == gr.1)] = 0
df$Group[which(df$Group == gr.2)] = 1
group.0 = df[df$Group == 0, ]
group.1 = df[df$Group == 1, ]
points = seq(1, 500)
model = curveFitting(formula = Count ~ Time, df = df, fit.method = "ssgaussian", points = points)

aametwally/OmicsLonDA documentation built on Feb. 10, 2020, 11:47 a.m.