w4mkmeans is a Planemo-built galaxy-tool-wrapper to wrap the stats::kmeans R package for clustering the dataMatrix produced by workflow4metabolomics (W4m) XCMS and CAMERA tools.
Using the intensities in the dataMatrix, w4mkmeans calculates the k-means clusters for samples or features and add them as columns to sampleMetadata or featureMetadata, respectively.
- If several, comma-separated K's are supplied, then one column is added for each K.
- Clustering is mutually exclusive, not hierarchical.
- For feature-clustering, each feature is assigned to a cluster such that the feature's response for all samples is closer to the mean of all features for that cluster than to the mean for any other cluster.
- For sample-clustering, each sample is assigned to a cluster such that the sample's response for all features is closer to the mean of all samples for that cluster than to the mean for any other cluster.