NBGOF
Goodness-of-Fit Tests and Model Diagnostics for Negative Binomial Regression of RNA sequencing Data
NBGOF is an R package for implementing goodness-of-fit (GOF) tests for negative binomial (NB) distributions and NB dispersion models, with applications in RNA-Seq data analysis. This package can be used to test the GOF of the NB2, NBP or Poisson regression models. It can also be used to test GOF for a variety of NB dispersion models in popular R/Bioconductor packages, including
-
NBP dispersion model in the NBPSeq package (NBP)
-
NBQ dispersion model in the NBPSeq package (NBQ)
-
NB common dispersion model in the edgeR package (Common)
-
NB genewise dispersion model in the edgeR package (Genewise)
-
NB trended (non-parametric) dispersion model in the edgeR package (Trended)
-
NB tagwise-common dispersion model in the edgeR package (Tagwise-Common)
-
NB tagwise-trended dispersion model in the edgeR package (Tagwise-Trend)
The methodologies are discussed in the manuscript Goodness-of-Fit Tests and Model Diagnostics for Negative Binomial Regression of RNA Sequencing Data (by Gu Mi, Yanming Di, and Daniel W. Schafer, PLOS ONE, 10(3)). The paper is freely available from this link. Functions in NBGOF have been used to generate all figures and tables displayed in the manuscript, plus some additional analysis tools for further investigations.
The NBGOF source package can be downloaded here.
Install
Installation from this repository requires the devtools package pre-installed. Run the following R command to install NBGOF:
devtools::install_github("gu-mi/NBGOF")To load the package into current R session, run
library(NBGOF)Two main functions for testing the adequacy of regression models and NB dispersion models are nb.gof.v and nb.gof.m, repectively. We provide two real datasets (arab and earthquake) to illustrate the use of this package.
We provide (Dropbox) links below to download R source codes and related supporting files for preparing the datasets and reproducing figures/tables in the manuscript. Some intermediate key results are also provided when necessary.
Figures
- Figure 1 -- The mean-dispersion plot with six fitted dispersion models (common, NBP, NBQ, trended, tagwise-common and tagwise-trend) for the Arabidopsis RNA-Seq dataset (19,623 genes from three biological samples in the mock treatment group): Download R files and results
- Figure 2 -- Empirical probability plots with GOF test p-values for evaluating NB2 and NBP model fits on the earthquake dataset: Download R files and results
- Figure 3 -- Empirical probability plots and GOF p-values for testing NB2 (top row) and NBP (bottom) on four simulated datasets with sample size = 45: Download R files and results
- Figure 4 -- Uniform QQ plots of individual GOF test p-values for the Arabidopsis dataset (based on a random sample of 1,000 genes from six experimental units in two experimental groups): Download R files and results
- SI Figure 1 -- Uniform QQ plots of individual GOF test p-values for the simulated ``NB2+noise'' dataset: Download R files and results
Tables
- Table 1 -- Type I error rates for 0.05-level NB2 and NBP GOF tests: Download R files
- Table 2 -- Rejection rates for 0.05-level NB2 and NBP GOF tests: Download R files
All results (figures/tables) in the submitted manuscript was produced using the following versions of R and its packages:
R version 3.2.2 (2015-08-14)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.11 (El Capitan)
locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
attached base packages:
[1] grid parallel stats graphics grDevices utils datasets methods base
other attached packages:
[1] NBGOF_0.2.0 ggplot2_1.0.1 doMC_1.3.3 iterators_1.0.7 foreach_1.4.2
loaded via a namespace (and not attached):
[1] Rcpp_0.12.1 NBPSeq_0.3.0 magrittr_1.5 edgeR_3.10.5 splines_3.2.2 MASS_7.3-44
[7] munsell_0.4.2 colorspace_1.2-6 stringr_1.0.0 plyr_1.8.3 tools_3.2.2 gtable_0.1.2
[13] digest_0.6.8 numDeriv_2014.2-1 reshape2_1.4.1 codetools_0.2-14 qvalue_2.0.0 labeling_0.3
[19] limma_3.24.15 stringi_0.5-5 compiler_3.2.2 scales_0.3.0 proto_0.3-10
If you have any questions, please do not hesitate to email the repository maintainer (Gu Mi) at neo.migu@gmail.com. Thank you for your interests in our research work.