Document Type
Article
Publication Date
12-1-2024
Original Citation
Grady S,
Peterson K,
Murray SA,
Baker E,
Langston M,
Chesler E.
A graph theoretical approach to experimental prioritization in genome-scale investigations. Mamm Genome. 2024;35(4):724-33.
Keywords
JMG, Animals, Mice, Humans, Algorithms, Gene Regulatory Networks, Computational Biology, Genome, Systems Biology, Genomics, Genome-Wide Association Study, Molecular Sequence Annotation
JAX Source
Mamm Genome. 2024;35(4):724-33.
ISSN
1432-1777
PMID
39191873
DOI
https://doi.org/10.1007/s00335-024-10066-z
Grant
This research has been supported in part by the National Institute of Child Health and Human Development under grant R01HD092653, and the NIH Knockout Mouse Program (KOMP) under grants U42 OD011185, U54 HG006332
Abstract
The goal of systems biology is to gain a network level understanding of how gene interactions influence biological states, and ultimately inform upon human disease. Given the scale and scope of systems biology studies, resource constraints often limit researchers when validating genome-wide phenomena and potentially lead to an incomplete understanding of the underlying mechanisms. Further, prioritization strategies are often biased towards known entities (e.g. previously studied genes/proteins with commercially available reagents), and other technical issues that limit experimental breadth. Here, heterogeneous biological information is modeled as an association graph to which a high-performance minimum dominating set solver is applied to maximize coverage across the graph, and thus increase the breadth of experimentation. First, we tested our model on retrieval of existing gene functional annotations and demonstrated that minimum dominating set returns more diverse terms when compared to other computational methods. Next, we utilized our heterogenous network and minimum dominating set solver to assist in the process of identifying understudied genes to be interrogated by the International Mouse Phenotyping Consortium. Using an unbiased algorithmic strategy, poorly studied genes are prioritized from the remaining thousands of genes yet to be characterized. This method is tunable and extensible with the potential to incorporate additional user-defined prioritizing information. The minimum dominating set approach can be applied to any biological network in order to identify a tractable subset of features to test experimentally or to assist in prioritizing candidate genes associated with human disease.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.