Role of Centrality in Network-Based Prioritization of Disease Genes

Abstract

Abstract. High-throughput molecular interaction data have been used effectively to prioritize candidate genes that are linked to a disease, based on the notion that the products of genes associated with similar diseases are likely to interact with each other heavily in a network of protein-protein interactions (PPIs). An important challenge for these applications, however, is the incomplete and noisy nature of PPI data. Random walk and network propagation based methods alleviate these problems to a certain extent, by considering indirect interactions and multiplicity of paths. However, as we demonstrate in this paper, such methods are likely to favor highly connected genes, making prioritization sensitive to the skewed degree distribution of PPI networks, as well as ascertainment bias in available interaction and disease association data. Here, we propose several statistical correction schemes that aim to account for the degree distribution of known disease and candidate genes. We show that, while the proposed schemes are very effective in detecting loosely connected disease genes that are missed by existing approaches, this improvement might come at the price of more false negatives for highly connected genes. Motivated by these results, we develop uniform prioritization methods that effectively integrate existing methods with the proposed statistical correction schemes. Comprehensive experimental results on the Online Mendelian Inheritance in Man (OMIM) database show that the resulting hybrid schemes outperform existing methods in prioritizing candidate disease genes.

Biomarker Identification for Prostate Cancer and Lymph Node Metastasis from Microarray Data and Protein Interaction Network Using Gene Prioritization Method

“…This model is applied using γ = 0.3 as probability of restart, as suggested by [22]. One disadvantage about the network based DGP is the use of noisy source data, therefore some steps are needed to filter out the source PINs in such a way that more relevant interactions are used in the core of the method.…”

Disease Gene Prioritization

“…One clear research opportunity is presented, and it is the combination of different network based approaches, by using local and global information. The work of (Erten & Koyutürk, 2010) shows promising results, aiming at the discovery of loosely connected genes using statistical correction schemes that help overcome the preference of straightforward method for genes with high centrality values; this …”

Vavien: An Algorithm for Prioritizing Candidate Disease Genes Based on Topological Similarity of Proteins in Interaction Networks

“…In other words, these methods work poorly in identifying loosely connected disease genes. Previous efforts reduce this bias to a certain extent by introducing several statistical correction schemes (Erten and Koyutürk, 2010). Motivated by these observations, we here investigate the effect of the bias introduced by degree distribution on the performance of different algorithms.…”

“…Information flow based algorithms are previously shown to be biased with respect to the degree of the target genes (Erten and Koyutürk, 2010). In other words, these methods work poorly in identifying loosely connected disease genes.…”

“…The performance here is measured in terms of the average rank of the target gene among 100 candidate genes, a lower value indicating better performance. Information flow with statistical correction: Based on the observation that the performance of information flow-based algorithms (including random walk with restarts and network propagation) depend on network degree, this algorithm applies statistical correction to the random walk-based association scores based on a reference model that takes into account the degree distribution of the PPI network (Erten and Koyutürk, 2010). Genomic Data Fusion (Endeavour): Endeavour is a comprehensive and flexible software package, which is freely available.…”