PoisMixMean function

Calculate the conditional per-cluster mean of each observation

This function is used to calculate the conditional per-cluster mean expression for all observations. This value corresponds to $\mu = (\mu_{ijlk}) = (\hat{w}_i \hat{\lambda}_{jk})$

for the PMM-I model and $\mu = (\mu_{ijlk}) = (\hat{w}_i s_{jl}\hat{\lambda}_{jk})$

for the PMM-II model.

PoisMixMean(y, g, conds, s, lambda)

Arguments

• y: (n x q) matrix of observed counts for n observations and q variables
• g: Number of clusters
• conds: Vector of length q defining the condition (treatment group) for each variable (column) in y
• s: Estimate of normalized per-variable library size
• lambda: (d x g) matrix containing the current estimate of lambda, where d is the number of conditions (treatment groups) and g is the number of clusters

Returns

A list of length g containing the (n x q) matrices of mean expression for all observations, conditioned on each of the g clusters

References

Rau, A., Maugis-Rabusseau, C., Martin-Magniette, M.-L., Celeux G. (2015). Co-expression analysis of high-throughput transcriptome sequencing data with Poisson mixture models. Bioinformatics, 31(9):1420-1427.

Rau, A., Celeux, G., Martin-Magniette, M.-L., Maugis-Rabusseau, C. (2011). Clustering high-throughput sequencing data with Poisson mixture models. Inria Research Report 7786. Available at https://inria.hal.science/inria-00638082.

Author(s)

Andrea Rau

See Also

PoisMixClus for Poisson mixture model estimation and model selection

Examples

set.seed(12345)

## Simulate data as shown in Rau et al. (2011)
## Library size setting "A", high cluster separation
## n = 200 observations

simulate <- PoisMixSim(n = 200, libsize = "A", separation = "high")
y <- simulate$y
conds <- simulate$conditions
s <- colSums(y) / sum(y)        ## TC estimate of lib size

## Run the PMM-II model for g = 3
## "TC" library size estimate, EM algorithm

run <- PoisMixClus(y, g = 3, norm = "TC", conds = conds) 
pi.est <- run$pi
lambda.est <- run$lambda

## Calculate the per-cluster mean for each observation
means <- PoisMixMean(y, g = 3, conds, s, lambda.est)