Identification of relevant substrates is essential for elucidation of in vivo functions of peptidases. The recent availability of the complete genome sequences of many eukaryotic organisms holds the promise of identifying specific peptidase substrates by systematic proteome analyses in combination with computer-based screening of genome databases. Currently available proteomics and bioinformatics tools are not sufficient for reliable endopeptidase substrate predictions. To address these shortcomings the bioinformatics tool 'PEPS' (Prediction of Endopeptidase Substrates) has been developed and is presented here. PEPS uses individual rule-based endopeptidase cleavage site scoring matrices (CSSM). The efficiency of PEPS in predicting putative caspase 3, cathepsin B and cathepsin L cleavage sites is demonstrated in comparison to established algorithms. Mortalin, a member of the heat shock protein family HSP70, was identified by PEPS as a putative cathepsin L substrate. Comparative proteome analyses of cathepsin L-deficient and wild-type mouse fibroblasts showed that mortalin is enriched in the absence of cathepsin L. These results indicate that CSSM/PEPS can correctly predict relevant peptidase substrates.

Data organization and data mining represents one of the main challenges for modern high throughput technologies in pharmaceutical chemistry and medical chemistry. The presented open source documentation and analysis system provides an integrated solution (tutorial, setup protocol, sources, executables) aimed at substituting the traditionally used lab-book. The data management solution provided incorporates detailed information about the processing of the gels and the experimental conditions used and includes basic data analysis facilities which can be easily extended. The sample database and User-Interface are available free of charge under the GNU license from http://webber.physik.uni-freiburg.de/~fallerd/tutorial.htm.