NIH VideoCast - Construction of Protein PTM Networks by Data Mining, Text Mining, and Ontology Integration: Application to Multi-Faceted Disease Analysis

Skip Navigation

Construction of Protein PTM Networks by Data Mining, Text Mining, and Ontology Integration: Application to Multi-Faceted Disease Analysis

Loading... Loading video...

291 Views

Air date:	Thursday, April 2, 2015, 10:00:00 AM Time displayed is Eastern Time, Washington DC Local
Views:	Total views: 291 (105 Live, 186 On-demand)
Category:	Proteomics
Runtime:	01:11:39
Closed captioning:	Check box to display CC outside of the video Note: You can drag the captioning window around and resize it
Description:	Perturbations in post-translational modifications (PTMs) and their downstream effects are recognized as key drivers of disease. We have developed iPTMnet, employing an integrative bioinformatics approach—combining text mining, data mining, and ontological representation to capture rich PTM information—for PTM network and disease discovery. Text mining tools are used for full-scale mining of PubMed abstracts and PMC Open Access articles to identify PTM information (kinase, substrate, and site) and phosphorylation-dependent protein-protein interactions (PPIs) in their biological context, including disease consequences. Experimentally observed PTMs, including high-throughput proteomic data from curated PTM databases, are incorporated. Ontologies are used for knowledge representation, particularly the Protein Ontology (PRO) for representation of PTM proteoforms and complexes. The web portal (http://proteininformationresource.org/iPTMnet) supports online search and visual analysis, including multiple-sequence alignment views for comparison of PTM forms across organisms and Cytoscape visualization of PTM enzyme-substrate and PPI relationships in PTM interaction networks. We are conducting use cases for PTM-disease discovery. First, we analyzed phosphorylation-dependent PPIs related to the PI3K/AKT/mTOR pathway, which is deregulated in many cancers and is targeted by therapeutic kinase inhibitors. We classified interactions as pro- and anti-oncogenic to indicate potential mechanisms of kinase-inhibitor resistance. Second, we made a knowledge map of beta-catenin PTM proteoforms to interpret patterns of mutations in PTM sites occurring in different types of cancer, revealing multiple mechanisms through which mutations in beta-catenin PTM sites contribute to cancer. Finally, we are exploring the phosphorylation of pyruvate dehydrogenase (PDHA1), a mediator of the Warburg effect in cancer whose phosphorylation levels span a 64-fold range in breast cancer patient samples from phosphoproteomic data CPTAC. We are investigating the levels and activity of PDHA1 kinases and phosphatases in the same samples using genomic, transcriptomic (mRNA and miRNA), and proteomic data to gain insight into the efficacy of existing anti-cancer drugs that target PDHA1 phosphorylation. For more information go to http://proteome.nih.gov

NLM Title:	Construction of protein PTM networks by data mining, text mining, and ontology integration : application to multi-faceted disease analysis / Dr. Cathy Wu.
Author:	Wu, Cathy H. National Institutes of Health (U.S.),
Publisher:
Abstract:	(CIT): Perturbations in post-translational modifications (PTMs) and their downstream effects are recognized as key drivers of disease. We have developed iPTMnet, employing an integrative bioinformatics approach--combining text mining, data mining, and ontological representation to capture rich PTM information--for PTM network and disease discovery. Text mining tools are used for full-scale mining of PubMed abstracts and PMC Open Access articles to identify PTM information (kinase, substrate, and site) and phosphorylation-dependent protein-protein interactions (PPIs) in their biological context, including disease consequences. Experimentally observed PTMs, including high-throughput proteomic data from curated PTM databases, are incorporated. Ontologies are used for knowledge representation, particularly the Protein Ontology (PRO) for representation of PTM proteoforms and complexes. The web portal (http://proteininformationresource.org/iPTMnet) supports online search and visual analysis, including multiple-sequence alignment views for comparison of PTM forms across organisms and Cytoscape visualization of PTM enzyme-substrate and PPI relationships in PTM interaction networks. We are conducting use cases for PTM-disease discovery. First, we analyzed phosphorylation-dependent PPIs related to the PI3K/AKT/mTOR pathway, which is deregulated in many cancers and is targeted by therapeutic kinase inhibitors. We classified interactions as pro- and anti-oncogenic to indicate potential mechanisms of kinase-inhibitor resistance. Second, we made a knowledge map of beta-catenin PTM proteoforms to interpret patterns of mutations in PTM sites occurring in different types of cancer, revealing multiple mechanisms through which mutations in beta-catenin PTM sites contribute to cancer. Finally, we are exploring the phosphorylation of pyruvate dehydrogenase (PDHA1), a mediator of the Warburg effect in cancer whose phosphorylation levels span a 64-fold range in breast cancer patient samples from phosphoproteomic data CPTAC. We are investigating the levels and activity of PDHA1 kinases and phosphatases in the same samples using genomic, transcriptomic (mRNA and miRNA), and proteomic data to gain insight into the efficacy of existing anti-cancer drugs that target PDHA1 phosphorylation.
Subjects:	Data Mining Databases, Protein Protein Interaction Mapping--methods Protein Processing, Post-Translational--physiology Proteins--metabolism Systems Biology
Publication Types:	Lecture Webcast

NLM Classification:	QU 55
NLM ID:	101658179
CIT Live ID:	16038
Permanent link:	https://videocast.nih.gov/watch=16038

Related Videos...