Dataset Information

Revealing missing parts of the interactome via link prediction.

ABSTRACT: Protein interaction networks (PINs) are often used to "learn" new biological function from their topology. Since current PINs are noisy, their computational de-noising via link prediction (LP) could improve the learning accuracy. LP uses the existing PIN topology to predict missing and spurious links. Many of existing LP methods rely on shared immediate neighborhoods of the nodes to be linked. As such, they have limitations. Thus, in order to comprehensively study what are the topological properties of nodes in PINs that dictate whether the nodes should be linked, we introduce novel sensitive LP measures that are expected to overcome the limitations of the existing methods. We systematically evaluate the new and existing LP measures by introducing "synthetic" noise into PINs and measuring how accurate the measures are in reconstructing the original PINs. Also, we use the LP measures to de-noise the original PINs, and we measure biological correctness of the de-noised PINs with respect to functional enrichment of the predicted interactions. Our main findings are: 1) LP measures that favor nodes which are both "topologically similar" and have large shared extended neighborhoods are superior; 2) using more network topology often though not always improves LP accuracy; and 3) LP improves biological correctness of the PINs, plus we validate a significant portion of the predicted interactions in independent, external PIN data sources. Ultimately, we are less focused on identifying a superior method but more on showing that LP improves biological correctness of PINs, which is its ultimate goal in computational biology. But we note that our new methods outperform each of the existing ones with respect to at least one evaluation criterion. Alarmingly, we find that the different criteria often disagree in identifying the best method(s), which has important implications for LP communities in any domain, including social networks.

SUBMITTER: Hulovatyy Y

PROVIDER: S-EPMC3940777 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Revealing missing parts of the interactome via link prediction.

Hulovatyy Yuriy Y Solava Ryan W RW Milenković Tijana T

PloS one 20140303 3

Protein interaction networks (PINs) are often used to "learn" new biological function from their topology. Since current PINs are noisy, their computational de-noising via link prediction (LP) could improve the learning accuracy. LP uses the existing PIN topology to predict missing and spurious links. Many of existing LP methods rely on shared immediate neighborhoods of the nodes to be linked. As such, they have limitations. Thus, in order to comprehensively study what are the topological proper ...[more]

PMID: 24594900

Similar Datasets

Project description:Degeneration of the intervertebral disc (IVD) is a frequent cause for back pain in humans and dogs. Link-N stabilizes proteoglycan aggregates in cartilaginous tissues and exerts growth factor-like effects. The human variant of Link-N facilitates IVD regeneration in several species in vitro by inducing Smad1 signaling, but it is not clear whether this is species specific. Dogs with IVD disease could possibly benefit from Link-N treatment, but Link-N has not been tested on canine IVD cells. If Link-N appears to be effective in canines, this would facilitate translation of Link-N into the clinic using the dog as an in vivo large animal model for human IVD degeneration.This study's objective was to determine the effect of the human and canine variant of Link-N and short (s) Link-N on canine chondrocyte-like cells (CLCs) and compare this to those on already studied species, i.e. human and bovine CLCs. Extracellular matrix (ECM) production was determined by measuring glycosaminoglycan (GAG) content and histological evaluation. Additionally, the micro-aggregates' DNA content was measured. Phosphorylated (p) Smad1 and -2 levels were determined using ELISA.Human (s)Link-N induced GAG deposition in human and bovine CLCs, as expected. In contrast, canine (s)Link-N did not affect ECM production in human CLCs, while it mainly induced collagen type I and II deposition in bovine CLCs. In canine CLCs, both canine and human (s)Link-N induced negligible GAG deposition. Surprisingly, human and canine (s)Link-N did not induce Smad signaling in human and bovine CLCs. Human and canine (s)Link-N only mildly increased pSmad1 and Smad2 levels in canine CLCs.Human and canine (s)Link-N exerted species-specific effects on CLCs from early degenerated IVDs. Both variants, however, lacked the potency as canine IVD regeneration agent. While these studies demonstrate the challenges of translational studies in large animal models, (s)Link-N still holds a regenerative potential for humans.

Dataset Information

Revealing missing parts of the interactome via link prediction.

Publications

Revealing missing parts of the interactome via link prediction.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets