fit_models: Linear regression of the null and full models

In edge: Extraction of Differential Gene Expression

Description

fit_models fits a model matrix to each gene by using the least squares method. Model fits can be either statistic type "odp" (optimal discovery procedure) or "lrt" (likelihood ratio test).

Usage

fit_models(object, stat.type = c("lrt", "odp"), weights = NULL)

## S4 method for signature 'deSet'
fit_models(object, stat.type = c("lrt", "odp"),
  weights = NULL)

Arguments

object	S4 object: deSet.
stat.type	character: type of statistic to be used. Either "lrt" or "odp". Default is "lrt".
weights	matrix: weights for each observation. Default is NULL.

Details

If "odp" method is implemented then the null model is removed from the full model (see Storey 2007). Otherwise, the statistic type has no affect on the model fit.

Value

deFit object

Note

fit_models does not have to be called by the user to use odp, lrt or kl_clust as it is an optional input and is implemented in the methods. The deFit object can be created by the user if a different statistical implementation is required.

Author(s)

John Storey

References

Storey JD. (2007) The optimal discovery procedure: A new approach to simultaneous significance testing. Journal of the Royal Statistical Society, Series B, 69: 347-368.

Storey JD, Dai JY, and Leek JT. (2007) The optimal discovery procedure for large-scale significance testing, with applications to comparative microarray experiments. Biostatistics, 8: 414-432.

Storey JD, Xiao W, Leek JT, Tompkins RG, and Davis RW. (2005) Significance analysis of time course microarray experiments. Proceedings of the National Academy of Sciences, 102: 12837-12842.

See Also

deFit, odp and lrt

Examples

# import data
library(splines)
data(kidney)
age <- kidney$age
sex <- kidney$sex
kidexpr <- kidney$kidexpr
cov <- data.frame(sex = sex, age = age)

# create models
null_model <- ~sex
full_model <- ~sex + ns(age, df = 4)

# create deSet object from data
de_obj <- build_models(data = kidexpr, cov = cov, null.model = null_model,
full.model = full_model)

# retrieve statistics from linear regression for each gene
fit_lrt <- fit_models(de_obj, stat.type = "lrt") # lrt method
fit_odp <- fit_models(de_obj, stat.type = "odp") # odp method

# summarize object
summary(fit_odp)

edge documentation built on Nov. 8, 2020, 6:48 p.m.