Conserved water molecules on protein-DNA interface

Water molecule in protein structure is considered as conserved one if there are water molecules in approximately the same position in independently solved structures of homologous proteins.

Conserved water molecules are detected by wLake program ( wLake takes superimposed structures on input and detects clusters of closely located water molecules from different member of the superimposition. Default threshold for distance between oxygens of two HOH molecules is set to 1.5Å, clusters that share a considerable number of water molecules are joined. See Aksianov et al.2008 ( for details.

Only those clusters that include at least one water molecule on protein – DNA interface are stored in NPIDB complex. The threshold of the number of waters in detected clusters is set to two.

Conserved water molecules on protein – DNA interface can be considered as significant players in protein – DNA interaction (Ramakrishnan, Rajagopalan, 2012,, Spirakis et al., 2007,, Grischchus, Ferreti, 2001, If a cluster contains also a water molecule from a structure of single protein (no DNA in PDB file) than this water can be considered as preimmobilized at protein surface and playing a role in DNA recognition.

Presentation of data on interfacial conserved waters

Links to clusters of water molecules are presented for SCOP families, that are endowed with superimposed structures. These families are marked by asterisk at page of SCOP domains in NPIDB.

A page of clusters in .txt format contains an information on water molecules, one water per line. Information includes:

  • Cluster number
  • PDB code and chain
  • Region of domain (res_start, res_end)
  • Model number (superimposed structures have model numbers 1, 2, …)
  • Water number (ResNum in PDB definition) and chain

  • Total number of water molecules in cluster
  • DNA chains identifiers

A page of clusters in RasmolScript .spt format contains a script for RasMol or Jmol, that defines each cluster as wat1, wat2, etc. Additionally join of all water clusters is defined as wat.

Jmol generator page for clusters (Coming soon!)

A page contains a table, rows are clusters of waters (wat1, wat2 etc.), columns are models (i.e. entries from a file with superposed domains). Mark "+" in a cell means that the water cluster (row) contains water molecule from the model (column).

To generate Jmol representation use check boxes to select clusters of waters and models to be represented and press Jmol.

Warrning It not recomended to visualize more than 10 (?) superposed models simultaneousely because of significant time for uploading to Jmol and reaction on mouse manipulations.

© NPIDB team 2003 - 2017