ContaMiner and ContaBase: Automated identification of unwantedly crystallized protein contaminants

Abstract

Solving the phase problem in protein X-ray crystallography relies heavily on the identity of the crystallized protein, especially when molecular replacement (MR) methods are used. Yet, it is not uncommon that a contaminant crystallizes instead of the protein of interest. Such contaminants may be proteins from the expression host organism, protein fusion tags or proteins added during the purification steps. Many contaminants co-purify easily, crystallize and give good diffraction data. Identification of contaminant crystals may take time, since the presence of the contaminant is unexpected, and its identity unknown. We have established a webserver (ContaMiner) and a contaminant database (ContaBase), both available at strube.cbrc.kaust.edu.sa/contaminer/, to allow fast MR-based screening of crystallographic data against currently 76 known contaminants from more than seven different expression systems [1]. Here we present the latest development of ContaMiner. Novel features include us eof UglyMol[2] for an interactive online visualization of electron density maps of MR solutions, the possibility for screening of 'custom contaminants', and features for improved speed and performances. We are currently using ContaMiner to scan all PDB entries with suspiciously poor refinement statistics, and will present the results obtained.