ClusterWaters: Cluster Conserved Waters

In vanddraabe: Identification and Statistical Analysis of Conserved Waters Near Proteins

Description

Cluster the conserved waters.

Usage

ClusterWaters(data, cutoff.cluster, cluster.method = "complete")

Arguments

data	The water oxygens' X, Y, and Z coordinates, B-values, and occupancy values.
cutoff.cluster	Numerical value provided by the user for the distance between water oxygen atoms to form a cluster; default: 2.4 Angstroms.
cluster.method	Method of clustering the waters; default is "complete". Any other method accepted by the hclust function is appropriate. The original method used by Sanschagrin and Kuhn is the complete linkage clustering method and is the default. Other options include "ward.D" (equivilant to the only Ward option in R versions 3.0.3 and earlier), "ward.D2" (implements Ward's 1963 criteria; see Murtagh and Legendre 2014), "single" (related to the minimal spanning tree method and adopts a "friend of friends" clustering method), along with "average" (= UPGMA), "mcquitty" (= WPGMA), "median" (= WPGMC) or "centroid" (= UPGMC). Due to size limitations with stats::hclust() – specifically the "size cannot be NA nor exceed 65536" – fastcluster::hclust() is being used because it is a complete replacement of stats::hclust(), is fast (compared to stats::hclust()), and is able to accommodate dissimilarity matrices with more than 2^16 (65,536) observations.

Details

Calculate the conserved waters using a collection of crystallographic protein structures.

Value

This function returns:

• h2o.clusters.raw: Initial waters with assigned cluster ID

• h2o.clusters.summary: Each cluster's:

  • cluster ID

  • number of waters

  • percent conservation

  • X, Y, and Z cooridinates

  • bound water environment measurements

  • mean distance between waters comprising the cluster

  • mean distance between waters comprising the cluster and the cluster's centroid

• h2o.occurrence: A table indicating the structures (PDBs) contributing to each cluster. This summary table includes the PDB structure's:

  • resolution

  • R-free value

  • occupancy (mean and standard deviation)

  • mobility (mean and standard deviation)

  • B-value (mean and standard deviation)

  • number of waters in each cluster

  • number of waters passing the mobility cutoff

  • number of waters passing the normalized B-value

  • number of waters passing both cutoff values

  • percentage of waters passing both cutoffs

  • number of clusters the structure contributes to

  • True/False table indicating if the protein structure contributed to the water cluster

• clustering.info: size and timing information

Author(s)

Emilio Xavier Esposito [e-mail]

References

Paul C Sanschagrin and Leslie A Kuhn. Cluster analysis of consensus water sites in thrombin and trypsin shows conservation between serine proteases and contributions to ligand specificity. Protein Science, 1998, 7 (10), pp 2054-2064.
DOI: 10.1002/pro.5560071002
PMID: 9792092
WatCH webpage

Fionn Murtagh and Pierre Legendre. Ward's Hierarchical Agglomerative Clustering Method: Which Algorithms Implement Ward's Criterion? Journal of Classification, 2014, 31, (3), pp 274-295.
DOI: 10.1007/s00357-014-9161-z

Daniel Müllner. fastcluster: Fast Hierarchical, Agglomerative Clustering Routines for R and Python. Journal of Statistical Software, 2013, 53 (9)
DOI: 10.18637/jss.v053.i09 fastcluster webpage

vanddraabe documentation built on June 8, 2019, 1:03 a.m.