Dataset Information

Predicting receptor-ligand pairs through kernel learning.

ABSTRACT:

Background

Regulation of cellular events is, often, initiated via extracellular signaling. Extracellular signaling occurs when a circulating ligand interacts with one or more membrane-bound receptors. Identification of receptor-ligand pairs is thus an important and specific form of PPI prediction.

Results

Given a set of disparate data sources (expression data, domain content, and phylogenetic profile) we seek to predict new receptor-ligand pairs. We create a combined kernel classifier and assess its performance with respect to the Database of Ligand-Receptor Partners (DLRP) 'golden standard' as well as the method proposed by Gertz et al. Among our findings, we discover that our predictions for the tgf? family accurately reconstruct over 76% of the supported edges (0.76 recall and 0.67 precision) of the receptor-ligand bipartite graph defined by the DLRP "golden standard". In addition, for the tgf? family, the combined kernel classifier is able to relatively improve upon the Gertz et al. work by a factor of approximately 1.5 when considering that our method has an F-measure of 0.71 while that of Gertz et al. has a value of 0.48.

Conclusions

The prediction of receptor-ligand pairings is a difficult and complex task. We have demonstrated that using kernel learning on multiple data sources provides a stronger alternative to the existing method in solving this task.

SUBMITTER: Iacucci E

PROVIDER: S-EPMC3199765 | biostudies-literature | 2011 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting receptor-ligand pairs through kernel learning.

Iacucci Ernesto E Ojeda Fabian F De Moor Bart B Moreau Yves Y

BMC bioinformatics 20110811

<h4>Background</h4>Regulation of cellular events is, often, initiated via extracellular signaling. Extracellular signaling occurs when a circulating ligand interacts with one or more membrane-bound receptors. Identification of receptor-ligand pairs is thus an important and specific form of PPI prediction.<h4>Results</h4>Given a set of disparate data sources (expression data, domain content, and phylogenetic profile) we seek to predict new receptor-ligand pairs. We create a combined kernel classi ...[more]

PMID: 21834994

Similar Datasets

Project description:Melanoma is a malignancy of melanocytes, responsible for a high percentage of skin cancer mortality. Ligand-Receptor pairs, a type of cellular communication, are essential for tumor genesis, growth, metastasis, and prognosis. Yet, the role of Ligand-Receptor pairs in melanoma has not been fully elucidated. Our research focused on the function of Ligand-Receptor pairs in melanoma prognosis. We screened 131 melanoma prognosis corresponded ligand-receptor pairs by analyzing the TCGA data of melanoma and the 2293 LR pairs retrieved from the connectomeDB2020 database. And further developed subtypes of melanoma according to the expression of these ligand-receptor pairs by Consensus Clustering. Then we using lasso cox regression and stepwise multivariate regression analysis established a ligand-receptor pairs-based scoring model for the evaluation of melanoma prognosis. Our study demonstrated that the ligand-receptor pairs are vital to the molecular heterogeneity of melanoma, and characterized three different melanoma ligand-receptor pairs subtypes. Among them, the C3 subtype showed a better prognosis, while the C1 subtype exhibited a low prognosis state. And our analysis then found out that this could be related to the differed activation and inhabitation of the cell cycle and immune-related pathways. Using lasso cox regression and stepwise multivariate regression analysis, we further identified 9 key ligand-receptor pairs and established a scoring model that effectively correlated with the prognosis, immune pathways, and therapy of melanoma, showing that the LR.score model was a trustworthy and independent biomarker for melanoma prognosis evaluation. In sum, we found that ligand-receptor pairs are significantly associated with the prognosis and therapy of melanoma. And our ligand-receptor-based scoring model showed potential for the evaluation of melanoma prognosis and immune therapy outcome prediction, which is crucial to the survival for the patients.

Project description:Bacterial surface ligands mediate interactions with the host cell during association that determines the specific outcome for the host-microbe association. The association begins with receptors on the host cell binding ligands on the microbial cell to form a partnership that initiates responses in both cells. Methods to determine the specific cognate partnerships are lacking. Determining these molecular interactions between the host and microbial surfaces are difficult, yet crucial in defining biologically important events that are triggered during association of the microbiome, and critical in defining the initiating signal from the host membrane that results in pathology or commensal association. In this study, we designed an approach to discover cognate host-microbe receptor/ligand pairs using a covalent cross-linking strategy with whole cells. Protein/protein cross-linking occurred when the interacting molecules were within 9-12 Å, allowing for identification of specific pairs of proteins from the host and microbe that define the molecular interaction during association. To validate the method three different bacteria with three previously known protein/protein partnerships were examined. The exact interactions were confirmed and led to discovery of additional partnerships that were not recognized as cognate partners, but were previously reported to be involved in bacterial interactions. Additionally, three unknown receptor/ligand partners were discovered and validated with in vitro infection assays by blocking the putative host receptor and deleting the bacterial ligand. Subsequently, Salmonella enterica sv. Typhimurium was cross-linked to differentiated colonic epithelial cells (caco-2) to discover four previously unknown host receptors bound to three previously undefined host ligands for Salmonella. This approach resulted in a priori discovery of previously unknown and biologically important molecules for host/microbe association that were casually reported to mediate bacterial invasion. The whole cell cross-linking approach promises to enable discovery of possible targets to modulate interaction of the microbiome with the host that are important in infection and commensalism, both of with initiate a host response.

Dataset Information

Predicting receptor-ligand pairs through kernel learning.

Background

Results

Conclusions

Publications

Predicting receptor-ligand pairs through kernel learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets