glmWeightedF: Zero-inflation adjusted statistical tests for assessing...

In drisso/zinbwave: Zero-Inflated Negative Binomial Model for RNA-Seq Data

Zero-inflation adjusted statistical tests for assessing differential expression.

Description

This function recycles an old version of the glmLRT method that allows an F-test with adjusted denominator degrees of freedom to account for the downweighting in the zero-inflation model.

Usage

glmWeightedF(
  glmfit,
  coef = ncol(glmfit$design),
  contrast = NULL,
  ZI = TRUE,
  independentFiltering = TRUE,
  filter = NULL
)

Arguments

glmfit	a DGEGLM-class object, usually output from glmFit.
coef	integer or character vector indicating which coefficients of the linear model are to be tested equal to zero. Values must be columns or column names of design. Defaults to the last coefficient. Ignored if contrast is specified.
contrast	numeric vector or matrix specifying one or more contrasts of the linear model coefficients to be tested equal to zero. Number of rows must equal to the number of columns of design. If specified, then takes precedence over coef.
ZI	logical, specifying whether the degrees of freedom in the statistical test should be adjusted according to the weights in the fit object to account for the downweighting. Defaults to TRUE and this option is highly recommended.
independentFiltering	logical, specifying whether independent filtering should be performed.
filter	vector of values to perform filtering on. Default is the mean of the fitted values from glmfit.

Details

When 'independentFiltering=TRUE' (default) an independent filtering step is applied prior to the multiple testing procedure, as described in great details in the 'DESeq2“ vignette. The values in the 'padjFilter' column refer to this procedure. They are identical to the 'FDR' values if the filtering step does not remove any gene, since the function uses the Benjamini-Hochberg correction by default. If the procedure filters some genes, the adjusted p-values will typically result in greater power to detect DE genes. The theory behind independent filtering is described in Bourgon et al. (2010).

Note

This function uses an adapted version of the glmLRT function that was originally written by Gordon Smyth, Davis McCarthy and Yunshun Chen as part of the edgeR package. Koen Van den Berge wrote code to adjust residual degree of freedoom and added the independent filtering step.

References

McCarthy, DJ, Chen, Y, Smyth, GK (2012). Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Research 40, 4288-4297. Bourgon, Richard, Robert Gentleman, and Wolfgang Huber (2010) Independent Filtering Increases Detection Power for High-Throughput Experiments. PNAS 107 (21): 9546-51.

See Also

glmLRT

drisso/zinbwave documentation built on March 18, 2024, 5:13 p.m.