acc: Auto Cross Covariance (ACC) for Generating Scales-Based...
In Rcpi: Molecular Informatics Toolkit for Compound-Protein Interaction in Drug Discovery

Description Usage Arguments Details Value Note Author(s) References See Also Examples

Auto Cross Covariance (ACC) for Generating Scales-Based Descriptors of the Same Length

1	acc(mat, lag)

`mat`	A `p * n` matrix. Each row represents one scale (total `p` scales), each column represents one amino acid position (total `n` amino acids).
`lag`	The lag parameter. Must be less than the amino acids.

This function calculates the auto covariance and auto cross covariance for generating scale-based descriptors of the same length.

A length lag * p^2 named vector, the element names are constructed by: the scales index (crossed scales index) and lag index.

To know more details about auto cross covariance, see the references.

Nan Xiao <https://nanx.me>

Wold, S., Jonsson, J., Sj\"orstr\"om, M., Sandberg, M., & R\"annar, S. (1993). DNA and peptide sequences and chemical processes multivariately modelled by principal component analysis and partial least-squares projections to latent structures. Analytica chimica acta, 277(2), 239–253.

Sj\"ostr\"om, M., R\"annar, S., & Wieslander, A. (1995). Polypeptide sequence property relationships in Escherichia coli based on auto cross covariances. Chemometrics and intelligent laboratory systems, 29(2), 295–305.

See extractPCMScales for generalized scales-based descriptors. For more details, see extractPCMDescScales and extractPCMPropScales.

p = 8    # p is the scales number
n = 200  # n is the amino acid number
lag = 7  # the lag paramter
mat = matrix(rnorm(p * n), nrow = p, ncol = n)
acc(mat, lag)