Paper Info
Title
Pathway Enrichment Analysis for Untargeted Metabolomics
Abstract
Metabolomics-based studies have provided critical insights across many applications and now offer researchers an opportunity to collect information about thousands of small molecules in-bulk through untargeted metabolomics. However, taking advantage of this new development requires accurate identification of metabolites and their biological significance in a given sample, which unfortunately remains difficult. Pathway enrichment is a powerful method that can aid in addressing those tasks, but existing techniques intended for gene enrichment analysis are not directly applicable to untargeted metabolomics. In this poster we address the following problem: given a network model of the biological sample and a likelihood score of observing metabolites (nodes) within the network, compute the enrichment of pathways within the network model. We approach this challenge as an optimization problem, where a solution is defined as a particular assignment of mass features to candidate metabolites. The method generates possible assignments of features to compounds using in silico fragmentation tools (e.g., MetFrag [1], CFM-ID [2], and CSI:FingerID [3]) and spectral database (e.g., MassBank [4]) and then attempts to iteratively improve a possible solution. By developing this method, we enable the use of pathway enrichment as an effective way of metabolite identification in untargeted metabolomics.
Year
DOI
Venue
2017
10.1145/3107411.3108233
BCB
Keywords
Field
DocType
Metabolomics,in silico fragmentation,metabolite annotation,pathway enrichment,mass spectrometry
Computer science,Metabolomics,Artificial intelligence,Bioinformatics,Gene Enrichment,Optimization problem,Network model,Machine learning,In silico
Conference
ISBN
Citations 
PageRank 
978-1-4503-4722-8
0
0.34
References 
Authors
1
3
Name
Order
Citations
PageRank
Vladimir Porokhin100.34
xinmeng li201.01
Soha Hassoun3535241.27