Cargando…

Comparison of large networks with sub-sampling strategies

Networks are routinely used to represent large data sets, making the comparison of networks a tantalizing research question in many areas. Techniques for such analysis vary from simply comparing network summary statistics to sophisticated but computationally expensive alignment-based approaches. Mos...

Descripción completa

Detalles Bibliográficos
Autores principales: Ali, Waqar, Wegner, Anatol E., Gaunt, Robert E., Deane, Charlotte M., Reinert, Gesine
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group 2016
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4933923/
https://www.ncbi.nlm.nih.gov/pubmed/27380992
http://dx.doi.org/10.1038/srep28955
_version_ 1782441250024062976
author Ali, Waqar
Wegner, Anatol E.
Gaunt, Robert E.
Deane, Charlotte M.
Reinert, Gesine
author_facet Ali, Waqar
Wegner, Anatol E.
Gaunt, Robert E.
Deane, Charlotte M.
Reinert, Gesine
author_sort Ali, Waqar
collection PubMed
description Networks are routinely used to represent large data sets, making the comparison of networks a tantalizing research question in many areas. Techniques for such analysis vary from simply comparing network summary statistics to sophisticated but computationally expensive alignment-based approaches. Most existing methods either do not generalize well to different types of networks or do not provide a quantitative similarity score between networks. In contrast, alignment-free topology based network similarity scores empower us to analyse large sets of networks containing different types and sizes of data. Netdis is such a score that defines network similarity through the counts of small sub-graphs in the local neighbourhood of all nodes. Here, we introduce a sub-sampling procedure based on neighbourhoods which links naturally with the framework of network comparisons through local neighbourhood comparisons. Our theoretical arguments justify basing the Netdis statistic on a sample of similar-sized neighbourhoods. Our tests on empirical and synthetic datasets indicate that often only 10% of the neighbourhoods of a network suffice for optimal performance, leading to a drastic reduction in computational requirements. The sampling procedure is applicable even when only a small sample of the network is known, and thus provides a novel tool for network comparison of very large and potentially incomplete datasets.
format Online
Article
Text
id pubmed-4933923
institution National Center for Biotechnology Information
language English
publishDate 2016
publisher Nature Publishing Group
record_format MEDLINE/PubMed
spelling pubmed-49339232016-07-08 Comparison of large networks with sub-sampling strategies Ali, Waqar Wegner, Anatol E. Gaunt, Robert E. Deane, Charlotte M. Reinert, Gesine Sci Rep Article Networks are routinely used to represent large data sets, making the comparison of networks a tantalizing research question in many areas. Techniques for such analysis vary from simply comparing network summary statistics to sophisticated but computationally expensive alignment-based approaches. Most existing methods either do not generalize well to different types of networks or do not provide a quantitative similarity score between networks. In contrast, alignment-free topology based network similarity scores empower us to analyse large sets of networks containing different types and sizes of data. Netdis is such a score that defines network similarity through the counts of small sub-graphs in the local neighbourhood of all nodes. Here, we introduce a sub-sampling procedure based on neighbourhoods which links naturally with the framework of network comparisons through local neighbourhood comparisons. Our theoretical arguments justify basing the Netdis statistic on a sample of similar-sized neighbourhoods. Our tests on empirical and synthetic datasets indicate that often only 10% of the neighbourhoods of a network suffice for optimal performance, leading to a drastic reduction in computational requirements. The sampling procedure is applicable even when only a small sample of the network is known, and thus provides a novel tool for network comparison of very large and potentially incomplete datasets. Nature Publishing Group 2016-07-06 /pmc/articles/PMC4933923/ /pubmed/27380992 http://dx.doi.org/10.1038/srep28955 Text en Copyright © 2016, Macmillan Publishers Limited http://creativecommons.org/licenses/by/4.0/ This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
spellingShingle Article
Ali, Waqar
Wegner, Anatol E.
Gaunt, Robert E.
Deane, Charlotte M.
Reinert, Gesine
Comparison of large networks with sub-sampling strategies
title Comparison of large networks with sub-sampling strategies
title_full Comparison of large networks with sub-sampling strategies
title_fullStr Comparison of large networks with sub-sampling strategies
title_full_unstemmed Comparison of large networks with sub-sampling strategies
title_short Comparison of large networks with sub-sampling strategies
title_sort comparison of large networks with sub-sampling strategies
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4933923/
https://www.ncbi.nlm.nih.gov/pubmed/27380992
http://dx.doi.org/10.1038/srep28955
work_keys_str_mv AT aliwaqar comparisonoflargenetworkswithsubsamplingstrategies
AT wegneranatole comparisonoflargenetworkswithsubsamplingstrategies
AT gauntroberte comparisonoflargenetworkswithsubsamplingstrategies
AT deanecharlottem comparisonoflargenetworkswithsubsamplingstrategies
AT reinertgesine comparisonoflargenetworkswithsubsamplingstrategies