Genome-wide functional association networks: background, data & state-of-the-art resources

Abstract

The vast amount of experimental data from recent advances in the field of high-throughput biology begs for integration into more complex data structures such as genome-wide functional association networks. Such networks have been used for elucidation of the interplay of intra-cellular molecules to make advances ranging from the basic science understanding of evolutionary processes to the more translational field of precision medicine. The allure of the field has resulted in rapid growth of the number of available network resources, each with unique attributes exploitable to answer different biological questions. Unfortunately, the high volume of network resources makes it impossible for the intended user to select an appropriate tool for their particular research question. The aim of this paper is to provide an overview of the underlying data and representative network resources as well as to mention methods of integration, allowing a customized approach to resource selection. Additionally, this report will provide a primer for researchers venturing into the field of network integration.

Keywords: Bayesian classification; functional association networks; network inference; protein–protein interactions.

Figure 1

Typical evidence types. The evidence typ es’ data are generated using different experiments or computational techniques for different levels of biological information, i.e. DNA (yellow), RNA (turquoise) or proteins (red). There are different techniques for evaluating the information content of each data type e.g. correlation measures such as MI and Pearson correlation for mRNA co-expression.

Figure 2

Example workflow to infer a functional association network. All supervised inference methods use a diverse collection of gold standard set. These sets represent known interactions between genes or protein. For example, proteins that physically interact or genes that are present in the same tissue or pathway. Once these homogeneous sets are defined, they can be applied separately as labels to an assemblage of input data spanning from co-expression over domain interaction to phylogenetic profiles. The labeled data are further used to train supervised methods and predict interactions between genes that have been not known so far. Each prediction is still specific to the gold standard used for labeling the data to begin with. To increase the power and coverage most methods summarize all networks in a ensemble network, incorporating all interaction types. Since the interaction loses its gold standard specific meaning it is newly defined as functional association.