2014/07/17 by Joseph Crawford, Yihan Sun, Crawford, Joseph +3 · 1 citation
Biochemistry, Genetics and Molecular Biology · #Bioinformatics and Genomic Networks #FOS: Biological sciences #Gene expression and cancer classification #Genomics and Phylogenetic Studies #Molecular Networks (q-bio.MN) #q-bio.MN
paper · pdf · doi:10.48550/arxiv.1407.4824
19 pages. 10 figures. Presented at the 2014 ISMB Conference, July 13-15, Boston, MA
arxiv created 2014/07/17 · openalex publication_date 2014/07/17 · arxiv updated 2014/07/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Biological network alignment identifies topologically and functionally conserved regions between networks of different species. It encompasses two algorithmic steps: node cost function (NCF), which measures similarities between nodes in different networks, and alignment strategy (AS), which uses these similarities to rapidly identify high-scoring alignments. Different methods use both different NCFs and different ASs. Thus, it is unclear whether the superiority of a method comes from its NCF, its AS, or both. We already showed on MI-GRAAL and IsoRankN that combining NCF of one method and AS of another method can lead to a new superior method. Here, we evaluate MI-GRAAL against newer GHOST to potentially further improve alignment quality. Also, we approach several important questions that have not been asked systematically thus far. First, we ask how much of the node similarity information in NCF should come from sequence data compared to topology data. Existing methods determine this more-less arbitrarily, which could affect the resulting alignment(s). Second, when topology is used in NCF, we ask how large the size of the neighborhoods of the compared nodes should be. Existing methods assume that larger neighborhood sizes are better. We find that MI-GRAAL's NCF is superior to GHOST's NCF, while the performance of the methods' ASs is data-dependent. Thus, the combination of MI-GRAAL's NCF and GHOST's AS could be a new superior method for certain data. Also, which amount of sequence information is used within NCF does not affect alignment quality, while the inclusion of topological information is crucial. Finally, larger neighborhood sizes are preferred, but often, it is the second largest size that is superior, and using this size would decrease computational complexity. Together, our results give several general recommendations for a fair evaluation of network alignment methods.