Discriminative topological features reveal biological network mechanisms
By Manuel Middendorf, Etay Ziv, Carter Adams, Jennifer Hom, Robin Koytcheff, Chaya Levovitz, Gregory Woods, Linda Chen, Chris H. Wiggins. 2004
3,951 citations recorded by the source corpus.
BackgroundRecent genomic and bioinformatic advances have motivated the development of numerous network models intending to describe graphs of biological, technological, and sociological origin. In most cases the success of a model has been evaluated by how well it reproduces a few key features of the real-world data, such as degree distributions, mean geodesic lengths, and clustering coefficients. Often pairs of models can reproduce these features with indistinguishable fidelity despite being generated by vastly different mechanisms. In such cases, these few target features are insufficient to distinguish which of the different models best describes real world networks of interest; moreover, it is not clear a priori that any of the presently-existing algorithms for network generation offers a predictive description of the networks inspiring them.ResultsWe present a method to assess systematically which of a set of proposed network generation algorithms gives the most accurate description of a given biological network. To derive discriminative classifiers, we construct a mapping from the set of all graphs to a high-dimensional (in principle infinite-dimensional) "word space". This map defines an input space for classification schemes which allow us to state unambiguously which models are most descriptive of a given network of interest. Our training sets include networks generated from 17 models either drawn from the literature or introduced in this work. We show that different duplication-mutation schemes best describe the E. coli genetic network, the S. cerevisiae protein interaction network, and the C. elegans neuronal network, out of a set of network models including a linear preferential attachment model and a small-world model.ConclusionsOur method is a first step towards systematizing network models and assessing their predictability, and we anticipate its usefulness for a number of communities.
Subfields: Aging, Geometry and Topology, Molecular Biology
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Molecular Biology
Citations in the corpus
- Simulation of the clinical and pathological manifestations of Coronavirus Disease 2019 (COVID-19) in golden Syrian hamster model: implications for disease pathogenesis and transmissibility (2020, methodology): “Missing loop residues in the SARS-CoV-2 spike protein RBD (PDB ID:6VSB) was modelled with I-TASSER.(13) The human ACE2 and SARS-CoV-spike complex structure model (PDB ID:2AJF) was used to model ACE2 and SARS-CoV-2-spike complex by superposition of SARS-CoV-2-RBD to SARS-CoV-RBD.”
- Computational Methods in Drug Discovery (2014, methodology): “Fragment-based methods like I-TASSER (Zhang, 2008) have performed well in CASP.”
- Using iterative fragment assembly and progressive sequence truncation to facilitate phasing and crystal structure determination of distantly related proteins. (2016, background): “Large-scale benchmarking data built on 500 nonredundant proteins (Zhang, 2008a) have shown a strong correlation between the C-score and TM-score values of I-TASSER models (Pearson correlation coefficient = 0.91).”
- TP53 copy number expansion is associated with the evolution of increased body size and an enhanced DNA damage response in elephants (2015, methodology): “To infer if derived residues in the TP53RTG12 dimerization interface might disrupt a physical interaction between TP53RTG12 and TP53 we generated a homology model of the TP53RTG12/TP53 dimer using I-TASSER/ModRefiner (Roy et al., 2010; Xu and Zhang, 2011; Zhang, 2008) and the crystal structure of th”
- Protein sequence comparison and fold recognition: progress and good-practice benchmarking. (2011, background): “To increase information content over the three secondary structure states, SAM-T2K and SP5 define finer alphabets of backbone structure [12,16,17].”
- Design of a specific peptide against phenolic glycolipid-1 from Mycobacterium leprae and its implications in leprosy bacilli entry (2022, methodology): “The peptide was mapped in the LAMA2 protein and modeled by using I-TASSER(24) and AlphaFold2 (https:// colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold2.ipynb).”