max planck institut
mpii logo Minerva of the Max Planck Society

DASMI Download

On this page you can find software libraries and documentation that allows you to extend the DASMI system to your needs.

Please note that it is not our intention to offer download options for all complete interaction datasets. We generally recommend visiting the websites of the original data providers, should you wish to download a complete interaction dataset for offline analysis (we provide links to all resources that DASMIweb uses on the data sources documentation page). Nevertheless, we provide most of the datasets that we temporarily cache at our institute for download. These caches and the associated download options will be removed, however, as soon as the original data providers offer data servers that DASMIweb can use directly.


1) Adding your own interaction data to DASMIweb

One way to make your interaction data or interaction confidence scores available to DASMIweb and other DASMI clients is to set up your own DASMI server. We provide a detailed tutorial (PDF, 937kb) that explains in detail how you can accomplish this with the Java DAS server Dazzle. A precompiled version of Dazzle with a number of example files can be downloaded here (ZIP, 16.609kb). If you encounter any problems with the tutorial, please contact us. Another server that can be used for providing interaction data is the Perl-written ProServer, developed at the Wellcome Trust Sanger Institute.


2) Creating your own DASMI client.

We provide a DASMI client library in case you want to use all the interaction data and interaction confidence scores that are available via DASMI servers in your own application. The source code of the Dasobert DAS client library that underlies DASMIweb is freely available here. A second DAS client library is Bio-Das-Lite, developed at the Wellcome Trust Sanger Institute.


3) Interaction datasets for offline analysis.

As described above, we here provide download options for most of the datasets that we temporarily cache at our institute. The plain text files contain binary interaction partners in two columns, one line per interaction. We do not offer download options for resources that are retrieved from external servers (BioGRID, IntAct, MINT, MPIDB) or where legal restrictions prevent us from providing the complete datasets (DIP, HiMAP, HiMAP-core, HPRD). Generally, we advice you to visit the website of the respective resources linked below for downloading fully-detailed complete datasets. If you are affiliated with one of the resources listed below and want us to remove the download option for your data, please contact us