Reconstructing Signaling Pathways Using Regular-Language Constrained Paths

Signaling pathways are widely studied in systems biology. Several databases catalog our knowledge of these pathways, including the proteins and interactions that comprise them. However, high-quality curation of this information is slow and painstaking. As a result, many interactions still lack annotation concerning the pathways they participate in. A natural question that arises is whether or not it is possible to automatically leverage existing annotations to identify new interactions for inclusion in a given pathway.

Here, we present RegLinker, an algorithm that achieves this purpose by computing multiple short paths from pathway receptors to transcription factors (TFs) within a background interaction network. The key idea underlying RegLinker is the use of regular-language constraints to control the number of non-pathway edges present in the computed paths. We systematically evaluate RegLinker and alternative approaches against a comprehensive set of 15 signaling pathways and demonstrate that RegLinker exhibits superior recovery of withheld pathway proteins and interactions. These results show the promise of our approach for prioritizing candidates for experimental study and the broader potential of automated analysis to attenuate difficulties of traditional manual inquiry.