RSPS-package: Provides estimates of power or sample size using empirical...

In RSPS: RNA-Seq Power Simulation

Description

The R package RSPS uses an efficient simulation algorithm to empirically determine statistical power and necessary sample size for RNA-Seq studies. The data are simulated from Poisson (no overdispersion) or Negative Binomial distribution (overdispersion). The package allows one to monitor the progress of the function when the power is being computed. There are two functions for providing plots of the estimated power for given sample size and estimated sample size to achieve desired power.

Details

Package:	RSPS
Type:	Package
Version:	1.0
Date:	2015-05-19
License:	GPL-2

Author(s)

Milan Bimali Maintainer: Milan Bimali <[e-mail]>

References

If necessary, users are referred to power and sample sizes computation found in introductory statistics textbooks.

See Also

rpoiss, rnbinom

Examples

power.poiss <- poiss.pow(n=c(5,10,15),lambda1=c(3),k=c(2,2.5,3),
alpha=0.05,seed = 20,numsim=100,monitor=TRUE)
power.poiss
power.plot(power.poiss,cutoff=c(0.8,0.9))
# Another example (takes longer to run)
#power.poiss <- poiss.pow(n=c(3,5,10,15),lambda1=c(3),k=c(2,2.5,3),
#alpha=0.001,seed = 20,numsim=500,monitor=TRUE)
#power.poiss
#power.plot(power.poiss,cutoff=c(0.8,0.9))

Example output

  |                                                                            
  |                                                                      |   0%
  |                                                                            
  |========                                                              |  11%
  |                                                                            
  |================                                                      |  22%
  |                                                                            
  |=======================                                               |  33%
  |                                                                            
  |===============================                                       |  44%
  |                                                                            
  |=======================================                               |  56%
  |                                                                            
  |===============================================                       |  67%
  |                                                                            
  |======================================================                |  78%
  |                                                                            
  |==============================================================        |  89%
  |                                                                            
  |======================================================================| 100%
   N Mean.Null Effect.Size Type.I.Error Power Std.Err
1  5         3         2.0         0.05  0.66   0.047
2 10         3         2.0         0.05  0.94   0.024
3 15         3         2.0         0.05  0.99   0.010
4  5         3         2.5         0.05  0.93   0.026
5 10         3         2.5         0.05  0.99   0.010
6 15         3         2.5         0.05  1.00   0.000
7  5         3         3.0         0.05  0.98   0.014
8 10         3         3.0         0.05  1.00   0.000
9 15         3         3.0         0.05  1.00   0.000

RSPS documentation built on May 1, 2019, 10:17 p.m.