gbetattest: Beta t-tests within groups

In NBBttest: Negative Binomial Beta t-Test

Beta t-tests within groups

Description

Beta t-tests are conducted within groups,genes,or libraries.

Usage

gbetattest(xx, W, nci, na, nb, level, padjust_methods,C=1.222, side)

Arguments

xx	a datasheet consisting of n columns and m rows. Columns contain information and count data columns n must be 1 or more and m must be over 100.
W	numeric value. It is omega estimated from null simulation.
nci	int numeric value indicating number of information columns.
na	int numeric value indicating number of replicates in condition a.
nb	int numeric value indicating number of replicates in condition b.
level	string value. It has 6 options: "isoform", "sgRNA", "RNA", "polyA.gene", "CRISPR.gene" and "splicing.gene".
padjust_methods	padjust.methods can choose one of "holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", TX, and "none" where "fdr" = "BH", "TX" is Tan and Xu's method (2015) with C=1.222 for adjusting p-value.
C	float numeric value for specifying a multiple procedure. C=0 tells mbetattest to perform single tests, C=1.222 tells mbetattest to perform BH correction of pvalues, C>1000 tells mbetattest to perform Bonferroni correction of pvalues.
side	string value for specifying one-tail test or two-tail test: side="up" for left-tail test, side="down" for right-tail test and side="both" for two-tail tests.

Details

Beta t-test will be conducted within a specified group or at a specified level. If level="RNA", then beta t-tests will be conducted within a whole library or the whole data. If level= "isoform", then data will be sparated in two parts: single-isoform and multi-isoform datasets. Single-isoform RNA indicates that there is only one RNA isform within a gene, while multi-isoform RNAs indicate that there are at least two RNA isoformswithin a gene. For single-isoforms, data are as a group and beta t-tests will be performed in the group. For the multi-isoforms, t-test will be performed within genes. If level="polyA.gene" or "CRISPR.gene", then t-test will be performed at gene level. If level="splicing.gene", then t-values and p-values will be selected from gene groups with the least p-values.

Value

return a list containing dataset, t-values, corrected p-values, rhos and w.

Author(s)

Yuan-De Tan [e-mail]

References

Baggerly KA, Deng L, Morris JS, Aldaz CM (2003) Differential expression in SAGE: accounting for normal between-library variation.
Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series, B, 575, 289-300. doi:10.1111/j.2517-6161.1995.tb02031.x, https://www.jstor.org/stable/2346101.
Benjamini, Y., and Yekutieli, D. (2001). The control of the false discovery rate in multiple testing under dependency. Annals of Statistics, 29, 1165 -1188. doi:10.1214/aos/1013699998.
Tan YD, Xu H. A general method for accurate estimation of false discovery rates in identification of differentially expressed genes. Bioinformatics. 2014 Jul 15;30(14):2018-25. doi:10.1093/bioinformatics/btu124. Epub 2014 Mar 14. PMID: 24632499.

See Also

betattest

Examples

data(jkttcell)
colnames(jkttcell)[3]<-"Gene"
res.isfo<-gbetattest(xx=jkttcell[1:100,], W=1, nci=7,
na=3, nb=3, level="isoform", padjust_methods="fdr",C=0,side="both")
	

NBBttest documentation built on May 30, 2022, 1:05 a.m.