lmerSeq.contrast: Function to summarize individual linear contrasts...

In stop-pre16/lmerSeq: Wrappers for Fitting Linear Mixed Models to RNA-Seq Data

Function to summarize individual linear contrasts coefficients

Description

Conducts t- or F-tests on linear contrasts of regression coefficients from fits done using lmerSeq.fit function. If the contrast matrix has only 1 row, a t-test is done. If the contrast matrix has more than 1 row, an F-test is done

Usage

lmerSeq.contrast(
  lmerSeq_results = NULL,
  contrast_mat = NULL,
  p_adj_method = "BH",
  ddf = "Satterthwaite",
  sort_results = T,
  include_singular = F
)

Arguments

lmerSeq_results	Results object from running lmerSeq.fit
contrast_mat	Numeric matrix representing the contrast to be tested. Matrix must have the same number of columns as the number of coefficients in the model. If the matrix has multiple rows, a simultaneous F-test will be done
p_adj_method	Method for adjusting for multiple comparisons (default is Benjamini-Hochberg). See p.adjust.methods
ddf	Method for computing degrees of freedom and t-statistics. Options are "Satterthwaite" and "Kenward-Roger"
sort_results	Should the results table be sorted by adjusted p-value?

Examples

data("expr_data")
vst_expr <- expr_example$vst_expr
sample_meta_data <- expr_example$sample_meta_data

##Only including 10 genes in the expression matrix
vst_expr <- vst_expr[1:10, ]

##  Fit the Model
fit_lmerSeq <- lmerSeq.fit(form = ~ group * time + (1|ids),
                           expr_mat = vst_expr,
                           sample_data = sample_meta_data,
                           REML = T)


##  1 dimensional contrast (t-test)
cont_mat1 <- rbind(c(0, 1, 0, 1)) # group diff. at followup
contrast_summary1 <- lmerSeq.contrast(lmerSeq_results = fit_lmerSeq,
                                      contrast_mat = cont_mat1,
                                      p_adj_method = 'BH',
                                      ddf = 'Satterthwaite',
                                      sort_results = T)
print(contrast_summary1)

##  multi-dimensional contrast (F-test)
cont_mat2 <- rbind(c(0, 1, 0, 0),
                   c(0, 0, 1, 0),
                   c(0, 0, 0, 1)) # simultaneous test of all coefficients
contrast_summary2 <- lmerSeq.contrast(lmerSeq_results = fit_lmerSeq,
                                      contrast_mat = cont_mat2,
                                      p_adj_method = 'BH',
                                      ddf = 'Satterthwaite',
                                      sort_results = T)
print(contrast_summary2)

stop-pre16/lmerSeq documentation built on July 27, 2022, 9:08 a.m.