List update processing (LUP)--solving the sequence database update problem

Sequence databases of today require frequent updating. Mirror procedures to copy incrementally updated databases as cumulative sets are the preferred method and can be implemented by straightforward scripting. However, limited bandwidth of networks and the increase of data require more powerful paradigms to reduce the workload reliably. We suggest the List Update Processing (LUP) principle. The system has been implemented on an experimental basis to update the Swiss EMBnet Node (BioComputing Basel, CH) with data from the European Bioinformatics Institute (EMBL Outstation, Hinxton Hall, UK). The results obtained from the prototype suggest to expand the system to several sites.