Querying Provenance along with External Domain Data Using Prolog

Abstract

Bioinformaticians have relied on computational simulations to run their biological experiments. This is due to the advantages offered by existing approaches, including tools to manage and run experiments, verify results and capture/analyze provenance data. Provenance is metadata that helps scientists to analyze in silico experiments, better understand their results, and reproduce them. However, provenance data is usually not enough. To improve the knowledge about the experiment, scientists often need to use domain-specific data available on external sources along with provenance data that is captured during the experiment execution. Although most of the existing tools provide mechanisms to capture and analyze provenance data, they do not offer means to enrich provenance with external domain data, or, when they do it, they do not have mechanisms to query provenance and domain data together in an effective way. In this article, we present an approach to analyzing provenance and domain data together using Prolog. Our goal is to improve provenance analysis. As a proof of concept, we present a case study of phylogenetic analysis (a biological experiment). Our approach, however, is designed to be generic and can be applied to other domains.