hmm_smooth: hmm_smooth

In protHMM: Protein Feature Extraction from Profile Hidden Markov Models

hmm_smooth

Description

This feature smooths the HMM matrix H by using sliding window of length sw to incorporate information from up and downstream residues into each row of the HMM matrix. Each HMM row r_i is made into the summation of r_{i-(sw/2)}+... r_i...+r_{i+(sw/2)}, for i = 1:L, where L is the number of rows in H. For rows such as the beginning and ending rows, 0 matrices of dimensions sw/2, 20 are appended to the original matrix H.

Usage

hmm_smooth(hmm, sw = 7)

Arguments

hmm	The name of a profile hidden markov model file.
sw	The size of the sliding window.

Value

A matrix of dimensions L \times 20.

References

Fang, C., Noguchi, T., & Yamana, H. (2013). SCPSSMpred: A General Sequence-based Method for Ligand-binding Site Prediction. IPSJ Transactions on Bioinformatics, 6(0), 35–42.

Examples

h<- hmm_smooth(system.file("extdata", "1DLHA2-7", package="protHMM"))

protHMM documentation built on July 9, 2023, 6:10 p.m.