Compute VPC CI based on parametric bootstrap for one or more features.
getBootCI(
CountMatrix,
Strains,
which.features,
num.boot,
method = "NB-fit",
alpha = 0.05,
optimizer = "nlminb"
)A \(G\times N\) count matrix. \(G\) is the number of features; \(N\) is the total number of samples.
A \(1\times N\) vector of strain labels corresponding to each sample.
A \(1\times k\) vector of select feature numbers for which CI is desired. \(k\leq G\).
Number of bootstraps.
Which method should be used, "CP-fit", "NB-fit" (default), or "VST". "VST" method bootstraps data under negative binomial mixed models.
A numerical value between 0 and 1, indicating the significance level of the CI. The CI will be \(100*(1-\alpha)\) percent CI. Default value is 0.05.
A character string that determines which optimization routine is to be used. It is only used for method = "CP-fit". Possible choices are "nlminb" (default), "L-BFGS-B", and "bobyqa".
A list of two objects. The first object is a \(k \times 2\) matrix containing the CI. The second object consists of a \(k \times\)num.boot matrix of all bootsrapped VPC values.