summary-ZlmFit-method: Summarize model features from a 'ZlmFit' object

In RGLab/MAST: Model-based Analysis of Single Cell Transcriptomics

Summarize model features from a ZlmFit object

Description

Returns a data.table with a special print method that shows the top 2 most significant genes by contrast. This data.table contains columns:

primerid

the gene

component

C=continuous, D=discrete, logFC=log fold change, S=combined using Stouffer's method, H=combined using hurdle method

contrast

the coefficient/contrast of interest

ci.hi

upper bound of confidence interval

ci.lo

lower bound of confidence interval

coef

point estimate

z

z score (coefficient divided by standard error of coefficient)

Pr(>Chisq)

likelihood ratio test p-value (only if doLRT=TRUE)

Some of these columns will contain NAs if they are not applicable for a particular component or contrast.

Usage

## S4 method for signature 'ZlmFit'
summary(
  object,
  logFC = TRUE,
  doLRT = FALSE,
  level = 0.95,
  parallel = FALSE,
  ...
)

Arguments

object	A ZlmFit object
logFC	If TRUE, calculate log-fold changes, or output from a call to getLogFC.
doLRT	if TRUE, calculate lrTests on each coefficient, or a character vector of such coefficients to consider.
level	what level of confidence coefficient to return. Defaults to 95 percent.
parallel	If TRUE and option(mc.cores)>1 then multiple cores will be used in fitting.
...	ignored

Value

data.table

See Also

print.summaryZlmFit

Examples

data(vbetaFA)
z <- zlm(~Stim.Condition, vbetaFA[1:5,])
zs <- summary(z)
names(zs)
print(zs)
##Select `datatable` copmonent to get normal print method
zs$datatable
## Can use parallel processing for LRT now
summary(z, doLRT = TRUE, parallel = TRUE)

RGLab/MAST documentation built on Sept. 30, 2023, 1:08 p.m.