Reasoning over biological networks using maximum satisfiability
CP'12 Proceedings of the 18th international conference on Principles and Practice of Constraint Programming
| Hi-index | 0.00 |
Today's molecular biology is confronted with enormous amounts of data, generated by new high-throughput technologies, along with an increasing number of biological models available over web repositories. This poses new challenges for bioinformatics to invent methods coping with incompleteness, heterogeneity, and mutual inconsistency of data and models. To this end, we built the library BioASP, providing a framework for analyzing biological data and models with Answer Set Programming (ASP). Due to the expressive modeling language, the inherent tolerance of incomplete knowledge, and efficient solving engines, ASP has proven to be an excellent tool for solving a variety of biological questions. The BioASP library implements methods for analyzing metabolic and gene regulatory networks, consistency checking, diagnosing, and repairing biological data and models. In particular, it allows for computing predictions and generating hypotheses about required expansions of biological models. To accomplish this, expert knowledge of both the biological application and the ASP paradigm needs to be combined. In fact, the functionalities provided by the BioASP library exploit technical know-how of modeling (biological) problems in ASP and gearing ASP solvers' parameters to them. Often, such best-practice technology is the result of an exhaustive series of tests. The BioASP library %, we gather % this knowledge integrates our practical experience and offers them via easy-to-use Python functions, thus enabling ASP non-experts to solve biological questions with ASP.