EBHMMNBMultiEM_2chain: Run EBSeqHMM model with a fixed expected FC

In lengning/EBSeqHMM: Bayesian analysis for identifying gene or isoform expression changes in ordered RNA-seq experiments

Description

Run EBSeqHMM model with a fixed expected FC

Usage

EBHMMNBMultiEM_2chain(Data,
	NgVector=NULL, Conditions, AllTran=NULL,
	AllPi0=NULL, Terms=NULL,
	sizeFactors, NumTranStage=c(3,2),PriorFC=2,
	StateNames=c("Up","Down"),homo=FALSE,
	UpdateRd=5, PIBound=.9, UpdatePI=FALSE,Print=FALSE,
	WithinCondR=TRUE,
	PenalizeLowMed=TRUE, PenalizeLowMedQt=.1,PenalizeLowMedVal=10)

Arguments

Data	input data, rows are genes and columns are samples
NgVector	Ng vector; NULL for gene level data
Conditions	A factor indicates the condition (time/spatial point) which each sample belongs to.
AllTran	initial values for transition matrices
AllPi0	initial values for starting probabilities
Terms	Terms
sizeFactors	a vector indicates library size factors
StateNames	names of the hidden states
NumTranStage	number of states in two chains
PriorFC	target FC for gridient change
homo	whether the chain is assumed to be homogenious
UpdateRd	max number of iteration
UpdatePI	whether update the mixture proportion of two chains
PIBound	upper bound of the mixture proportion of the two chains
Print	Whether print the elapsed-time while running the test.
WithinCondR	By defining WithinCondR=TRUE, estimation of r's are obtained within each condition. (WithinCondR=FALSE is not suggested here)
PenalizeLowMed,PenalizeLowMedQt,PenalizeLowMedVal	Transcripts with median quantile < = PenalizeLowMedQt will be penalized

Details

EBHMMNBMultiEM_2chain() function implements the EBSeqHMM model to perform statistical analysis in an RNA-seq experiment with ordered conditions. EBHMMNBMultiEM_2chain() calls EBHMMNBfunForMulti() function to perform Balm-Welch algorithm that estimates the starting probabilities and transition probabilities. Here the emission distribution of each gene is assumed to be a Beta-Negative Binomial distribution with parameters (r_g, alpha, beta) , in which alpha and beta are shared by all the genes and r_g is gene specific. If not specified, r_g, alpha and beta will be estimated using method of moments. For isoform data, we assume isoforms from the same Ig group share the same beta^Ig. alpha is shared by all the isoforms and r_gi is isoform specific. The user also needs to specify an expected FC. Function EBSeqHMMTest() runs several models with varying FCs and returns the model with maximum likelihood.

Value

Pi0Out: estimated starting probabilities of each iteration.

TranOut: estimated transition probabilities of each iteration.

Pi: estimated probability of being each chain.

Alpha, Beta: estimated alpha and beta(s).

LLSum: log likelihood of the model.

QList: estimated q's.

MgAllPP: marginal PP for all paths.

MgAllMAPChar: Most likely path based on MgAllPP.

MgAllMaxVal: Highest PP based on MgAllPP.

PPMatW: Posterior probabilities of being each of the chains.

Author(s)

Ning Leng

Examples

data(GeneExampleData)
CondVector <- rep(paste("t",1:5,sep=""),each=3)
Conditions <- factor(CondVector, levels=c("t1","t2","t3","t4","t5"))
Sizes <- MedianNorm(GeneExampleData)
tmp <- EBHMMNBMultiEM_2chain(Data=GeneExampleData,sizeFactors=Sizes, Conditions=Conditions,
          UpdateRd=2)

lengning/EBSeqHMM documentation built on May 21, 2019, 4:02 a.m.