A Critical Examination of Complex Network File Formats for Bioinformatics Data Sources

Abstract

Recent work has demonstrated the effectiveness of using complex networks for studying complex biological interactions. The process often involves conversion of biological data such as gene expressions or biochemical interactions into one of numerous complex network file formats. However, the exact selection of an appropriate file format for a particular dataset often poses a non-trivial decision problem; biologists are non-specialist end-users and find it difficult to select a particular format for data storage and conversion. In this paper, we propose a solution to this problem of the selection of a suitable network format by means of a critical evaluation based on performance analysis of empirical experiments on biological data sets of different sizes. Experimental results substantiate the hypothesis of being extra careful in the selection of a suitable complex network format based primarily on the size of the biological dataset.