Human-like layout algorithms for signalling hypergraphs: outlining requirements.

The use of signalling pathway hypergraphs represented as process description diagrams is steadily becoming more pervasive in the field of biology. This makes ever more evident the necessity for an effective automated layout that can replicate high-quality manually drawn diagrams. The complexity and idiosyncrasies of these diagrams, as well as the specific tasks the end users perform with them, mean that a layout must meet many requirements beyond the simple metrics used in existing automated computational approaches. In this paper we outline these requirements, examine existing ones and describe new ones. We demonstrate state-of-the-art layout techniques enhanced with novel functionalities to meet part of the requirements. For comparatively small signalling pathways our enhanced algorithm provides results close to manually drawn layouts. In addition, we suggest technical approaches that may be suited for fulfilling the identified requirements currently not covered.

Machine Learning to Support the Presentation of Complex Pathway Graphs.

Visualization of biological mechanisms by means of pathway graphs is necessary to better understand the often complex underlying system. Manual layout of such pathways or maps of knowledge is a difficult and time consuming process. Node duplication is a technique that makes layouts with improved readability possible by reducing edge crossings and shortening edge lengths in drawn diagrams. In this article, we propose an approach using Machine Learning (ML) to facilitate parts of this task by training a Support Vector Machine (SVM) with actions taken during manual biocuration. Our training input is a series of incremental snapshots of a diagram describing mechanisms of a disease, progressively curated by a human expert employing node duplication in the process. As a test of the trained SVM models, they are applied to a single large instance and 25 medium-sized instances of hand-curated biological pathways. Finally, in a user validation study, we compare the model predictions to the outcome of a node duplication questionnaire answered by users of biological pathways with varying experience. We successfully predicted nodes for duplication and emulated human choices, demonstrating that our approach can effectively learn human-like node duplication preferences to support curation of pathway diagrams in various contexts.