First derivative of beta minus-log-pdf with respect to first parameter

Description

First derivative of -log(beta pdf) wrt a (first) parameter

Usage

diffBeta1(x,a,b)

Arguments

Details

First derivative of beta minus-log-pdf with respect to first parameter; used for maximum likelihood estimation, not typically called by user.

Value

first derivative with respect to a (first) parameter

Author(s)

E. Andres Houseman

See Also

diffBeta2,scoreOxBS

First derivative of beta minus-log-pdf with respect to second parameter

Description

First derivative of -log(beta pdf) wrt b (second) parameter

Usage

diffBeta2(x,a,b)

Arguments

Details

First derivative of beta minus-log-pdf with respect to second parameter; used for maximum likelihood estimation, not typically called by user.

Value

first derivative with respect to b (second) parameter

Author(s)

E. Andres Houseman

See Also

diffBeta1,scoreOxBS

Sample Data: Methylation (red) signals from conventional bisulfite conversion.

Description

Matrix of signal intensitities corresponding to 30 specimens and 30 CpGs.

Usage

exampleMethBS

Format

30 x 30 matrix (CpGs x Specimens)

Sample Data: Methylation (red) signals from oxy-bisulfite conversion.

Description

Matrix of signal intensitities corresponding to 30 specimens and 30 CpGs.

Usage

exampleMethOxBS

Format

30 x 30 matrix (CpGs x Specimens)

Sample Data: Unmethylated (green) signals from conventional bisulfite conversion.

Description

Matrix of signal intensitities corresponding to 30 specimens and 30 CpGs.

Usage

exampleUnmethBS

Format

30 x 30 matrix (CpGs x Specimens)

Sample Data: Unmethylated (green) signals from oxy-bisulfite conversion.

Description

Matrix of signal intensitities corresponding to 30 specimens and 30 CpGs.

Usage

exampleUnmethOxBS

Format

30 x 30 matrix (CpGs x Specimens)

Fit one OxyBS result

Description

Uses maximum likelihood to estimate (C,5mC,5hmC) for one CpG and one specimen

Usage

fitOneOxBS(betaBS, betaOxBS, signalBS, signalOxBS, eps=1E-5)

Arguments

Details

Uses maximum likelihood to estimate (C,5mC,5hmC) for one CpG and one specimen; not typically called by user.

Value

(C,5mC,5hmC) for one CpG and one specimen.

Author(s)

E. Andres Houseman

See Also

fitOxBS

Fit OxyBS for one specimen

Description

Uses maximum likelihood to estimate (C,5mC,5hmC) vectors for one specimen

Usage

fitOxBS(betaBS, betaOxBS, signalBS, signalOxBS, eps=1E-5)

Arguments

Details

Uses maximum likelihood to estimate (C,5mC,5hmC) one specimen (many CpGs).

Value

matrix of (C,5mC,5hmC) values (each row corresponds to a separate CpG).

Author(s)

E. Andres Houseman

Examples

## Not run: 
data(OxyBSSampleData)

nSpecimens <- 30
nCpGs <- 30

# Calculate Total Signals
signalBS <- exampleMethBS+exampleUnmethBS
signalOxBS <- exampleMethOxBS+exampleUnmethOxBS

# Calculate Beta Values
betaBS <- exampleMethBS/signalBS
betaOxBS <- exampleMethOxBS/signalOxBS

# Create container for results
MethOxy <- array(NA,dim=c(nCpGs,nSpecimens,3))
dimnames(MethOxy) <- list(
  rownames(exampleMethBS)[1:nCpGs],
  colnames(exampleMethBS)[1:nSpecimens],
  c("C","5mC","5hmC"))

# Process results (one array at a time)
for(i in 1:nSpecimens){
  MethOxy[,i,] <- fitOxBS(betaBS[,i],betaOxBS[,i],signalBS[,i],signalOxBS[,i])
}

# Check that results sum to one
table(apply(MethOxy,1:2,sum))

# First specimen
MethOxy[,1,]

# Ranges
range(MethOxy[,,1])
range(MethOxy[,,2])
range(MethOxy[,,3])

## End(Not run)

Likelihood function for C/5mC/5hmC likelihood estimator

Description

Likelihood function for C/5mC/5hmC likelihood estimator

Usage

likeOxBS(theta, betaBS, betaOxBS, signalBS, signalOxBS)

Arguments

Details

Likelihood function for C/5mC/5hmC likelihood estimator; used for maximum likelihood estimation, not typically called by user.

Value

likelihood for C/5mC/5hmC likelihood

Author(s)

E. Andres Houseman

See Also

fitOneOxBS

Score function for C/5mC/5hmC likelihood estimator

Description

Score function for C/5mC/5hmC likelihood estimator

Usage

scoreOxBS(theta, betaBS, betaOxBS, signalBS, signalOxBS)

Arguments

Details

Score function for C/5mC/5hmC likelihood estimator; used for maximum likelihood estimation, not typically called by user.

Value

score vector for C/5mC/5hmC likelihood

Author(s)

E. Andres Houseman

See Also

fitOneOxBS