Inter-cellular communication is central to immune system functionality. Though these complex signaling interactions are highly studied, the high system complexity, together with tremendous volume of accumulated knowledge, challenge human capability of reasoning over immune data. To address the deluge of knowledge and make it accessible for system-level reasoning, we built immuneXpresso, a Text Mining engine, that automatically extracts a global high-resolution directional cell-cytokine interaction network from PubMed abstracts.

The knowledgebase contains both interactions and separate mentions of cells or cytokines in context of thousands of diseases. As an experimental feature, interaction polarity (called "sentiment" interchangeably) indicating its positive, negative or neutral effect, as well as the cellular function involved are included. This network was already used to explore current knowledge, predict previously unappreciated cell-cytokine interactions or cytokine-disease associations, build immune-based global disease classification and assist data interpretation.

For each record, all evidence sentences are presented. The knowledgebase is heavily filtered, using manual curation and machine learning, to achieve high precision. Don't let paucity of evidence (possible for rare interactions) or misinterpretation of individual sentences to let you down - the real power stems from standardization, structuring and breadth of the global network.

We have lots of plans on expanding the data and improving the content quality, now making first steps towards transforming immunology to systemized, model-based science, a true "systems immunology'!

Start with specifying your cells and/or cytokines of interest within the search box on left (or leave it empty to see all), optionally limit by disease context, interaction or article features, click Search immuneXpresso and start connecting the dots!