Abstract
In my research I work with mathematical models of metabolism, the collection of all chemical reactions needed for cellular growth. The most popular models in this area are genome-scale models, which consist of equations representing all reactions for a given organism. These models are typically huge (>1000 equations), which leads to two problematic questions: 1) If a new model gets published, how can we quickly tell if it has good quality? and 2) if I find new information to update a model, how can I tell if the changes will be beneficial? To address these questions, in this paper a big part of the metabolic community agreed on a set of tests that can be done on a GEM to determine quality, mainly in terms of annotation and consistency. These tests were then packaged as an automatic tool (memote), ready to go to aid reviewers when assessing a new model, and model developers to conveniently track quality improvements as they add/fix parts of their models. I’m sure anyone that has ever used one of these models will appreciate this tool, I know I do.
