The NCutYX package includes functions for clustering genomic data using graph theory. Each function in this package is a variation on the NCut measure used to cluster vertices in a graph. The running theme is to use data sets from different sources and types to improve the clustering results.
The Normalized Cut (NCut) clusters the columns of Y into K groups using the NCut graph measure. Builds a similarity matrix for the columns of Y and clusters them into K groups based on the NCut graph measure. Correlation, Euclidean and Gaussian distances can be used to construct the similarity matrix. The NCut measure is minimized using the cross entropy method, a Monte Carlo optimization technique.
The Assisted NCut (ANcut) clusters the columns of a data set Y into K groups with the help of an external data set X, which is associated linearly with Y.
This example shows how to use the muncut function. MuNCut clusters the columns of data from 3 different sources. It clusters the columns of Z, Y and X into K clusters by representing each data type as one network layer. It represents the Z layer depending on Y, and the Y layer depending on X. Elastic net can be used before the clustering procedure by using the predictions of Z and Y instead of the actual values to improve the cluster results. The function muncut will output K clusters of columns of Z, Y and X.
The Penalized Weighted NCut (PWNCut) clusters the columns of X into K clusters by giving a weighted cluster membership while shrinking weights towards each other.
The Multilayer Biclustering NCut (MLBNCut) clusters the columns and the rows simultaneously of data from 3 different sources. It clusters the columns of Z,Y and X into K clusters and the samples into R clusters by representing each data type as one network layer. It represents the Z layer depending on Y, and the Y layer depending on X. This function will output K clusters of columns of Z, Y and X and R clusters of the samples.
The Assisted Weighted NCut builds the similarity matrices for the rows of X and an assisted dataset Z. Clusters them into K groups while conducting feature selection based on the AWNCut method.