Conference Paper

Homomorphisms of Multisource Trees into Networks with Applications to Metabolic Pathways

Georgia State Univ., Atlanta
DOI: 10.1109/BIBE.2007.4375587 Conference: Bioinformatics and Bioengineering, 2007. BIBE 2007. Proceedings of the 7th IEEE International Conference on
Source: IEEE Xplore

ABSTRACT Network mapping is a convenient tool for comparing and exploring biological networks; it can be used for predicting unknown pathways, fast and meaningful searching of databases, and potentially establishing evolutionary relations. Unfortunately, existing tools for mapping paths into general networks (PathBlast) or trees into tree networks allowing gaps (MetaPathwayHunter) cannot handle large query pathways or complex networks. In this paper we consider homomorphisms, i.e., mappings allowing to map different enzymes from the query pathway into the same enzyme from the networks. Homomorphisms are more general than homeomorphism (allowing gaps) and easier to handle algorithmically. Our dynamic programming algorithm efficiently finds the minimum cost homomorphism from a multisource tree to directed acyclic graphs as well as general networks. We have performed pairwise mapping of all pathways for four organisms (E. coli, S. cerevisiae, B. subtilis and T. thermophilus species) and found a reasonably large set of statistically significant pathway similarities. Further analysis of our mappings identifies conserved pathways across examined species and indicates potential pathway holes in existing pathway descriptions.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The network-mapping tool integrated with protein database search can be used for filling pathway holes. A metabolic pathway under consideration (pattern) is mapped into a known metabolic pathway (text), to find pathway holes. Enzymes that do not show up in the pattern may be a hole in the pattern pathway or an indication of alternative pattern pathway. We present a data-mining framework for filling holes in the pattern metabolic pathway based on protein function, prosite scan and protein sequence homology. Using this framework we suggest several fillings found with the same EC notation, with group neighbors (enzymes with same EC number in first three positions, different in the fourth position), and instances where the function of an enzyme has been taken up by the left or right neighboring enzyme in the pathway. The percentile scores are better when closely related organisms are mapped as compared to mapping distantly related organisms.
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this article we address two pattern matching problems which have important applications to bioinformatics. First we address the topology-free network query problem : Given a set of labels L , a multiset P of labels from L , a graph H=(VH,EH)H=(VH,EH) and a function LabelH:VH→2LLabelH:VH→2L, we need to find a subtree S of H which is an occurrence of P . We provide a parameterized algorithm with parameter k=|P|k=|P| that runs in time O⁎(2k)O⁎(2k) and whose space complexity is polynomial. We also consider three variants of this problem. Then we address the alignment network query problem: Given two labeled graphs P and H, we need to find a subgraph S of H whose alignment with P is the best among all such subgraphs. We present two algorithms for cases in which P and H belong to certain families of DAGs. Their running times are polynomial and they are less restrictive than algorithms that are available today for alignment network queries. Topology-free and alignment network queries provide means to study the function and evolution of biological networks, and today, with the increasing amount of knowledge regarding biological networks, they are extremely relevant.
    Journal of Discrete Algorithms 07/2014; 27. DOI:10.1016/j.jda.2014.03.002
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Network mappings has been commonly applied on it to detect con- served subnetworks across species in biological network. Our previous work fo- cused on finding best mapping of pattern graph to text graph in which vertices in the image of pattern graph are allowed to be deleted (1),(2). Arising from the requirement of allowing deleting pattern vertices in mapping, we reformulate the graph comparison problem to be the one of a homo-home morphism mapping among arbitrary graphs and further analyze and compare our work with previous ones.

Full-text (3 Sources)

Available from
May 31, 2014