Multi-layer sequential network analysis improves protein 3D structural classification.

Protein structural classification (PSC) is a supervised problem of assigning proteins into pre-defined structural (e.g., CATH or SCOPe) classes based on the proteins' sequence or 3D structural features. We recently proposed PSC approaches that model protein 3D structures as protein structure networks (PSNs) and analyze PSN-based protein features, which performed better than or comparable to state-of-the-art sequence or other 3D structure-based PSC approaches. However, existing PSN-based PSC approaches model the whole 3D structure of a protein as a static (i.e., single-layer) PSN. Because folding of a protein is a dynamic process, where some parts (i.e., sub-structures) of a protein fold before others, modeling the 3D structure of a protein as a PSN that captures the sub-structures might further help improve the existing PSC performance. Here, we propose to model 3D structures of proteins as multi-layer sequential PSNs that approximate 3D sub-structures of proteins, with the hypothesis that this will improve upon the current state-of-the-art PSC approaches that are based on single-layer PSNs (and thus upon the existing state-of-the-art sequence and other 3D structural approaches). Indeed, we confirm this on 72 datasets spanning ~44 000 CATH and SCOPe protein domains.

Modeling multi-scale data via a network of networks.

MOTIVATION: Prediction of node and graph labels are prominent network science tasks. Data analyzed in these tasks are sometimes related: entities represented by nodes in a higher-level (higher scale) network can themselves be modeled as networks at a lower level. We argue that systems involving such entities should be integrated with a 'network of networks' (NoNs) representation. Then, we ask whether entity label prediction using multi-level NoN data via our proposed approaches is more accurate than using each of single-level node and graph data alone, i.e. than traditional node label prediction on the higher-level network and graph label prediction on the lower-level networks. To obtain data, we develop the first synthetic NoN generator and construct a real biological NoN. We evaluate accuracy of considered approaches when predicting artificial labels from the synthetic NoNs and proteins' functions from the biological NoN. RESULTS: For the synthetic NoNs, our NoN approaches outperform or are as good as node- and network-level ones depending on the NoN properties. For the biological NoN, our NoN approaches outperform the single-level approaches for just under half of the protein functions, and for 30% of the functions, only our NoN approaches make meaningful predictions, while node- and network-level ones achieve random accuracy. So, NoN-based data integration is important. AVAILABILITY AND IMPLEMENTATION: The software and data are available at https://nd.edu/~cone/NoNs. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Inference of a Dynamic Aging-related Biological Subnetwork via Network Propagation.

Gene expression (GE)data capture valuable condition-specific information ("condition" can mean a biological process, disease stage, age, patient, etc.)However, GE analyses ignore physical interactions between gene products, i.e., proteins. Because proteins function by interacting with each other, and because biological networks (BNs)capture these interactions, BN analyses are promising. However, current BN data fail to capture condition-specific information. Recently, GE and BN data have been integrated using network propagation (NP)to infer condition-specific BNs. However, existing NP-based studies result in a static condition-specific subnetwork, even though cellular processes are dynamic. A dynamic process of our interest is human aging. We use prominent existing NP methods in a new task of inferring a dynamic rather than static condition-specific (aging-related)subnetwork. Then, we study evolution of network structure with age - we identify proteins whose network positions significantly change with age and predict them as new aging-related candidates. We validate the predictions via e.g., functional enrichment analyses and literature search. Dynamic network inference via NP yields higher prediction quality than the only existing method for inferring a dynamic aging-related BN, which does not use NP. Our data and code are available at https://nd.edu/~cone/dynetinf.

CHARMING: Harmonizing synonymous codon usage to replicate a desired codon usage pattern.

There is a growing appreciation that synonymous codon usage, although historically regarded as phenotypically silent, can instead alter a wide range of mechanisms related to functional protein production, a term we use here to describe the net effect of transcription (mRNA synthesis), mRNA half-life, translation (protein synthesis) and the probability of a protein folding correctly to its active, functional structure. In particular, recent discoveries have highlighted the important role that sub-optimal codons can play in modifying co-translational protein folding. These results have drawn increased attention to the patterns of synonymous codon usage within coding sequences, particularly in light of the discovery that these patterns can be conserved across evolution for homologous proteins. Because synonymous codon usage differs between organisms, for heterologous gene expression it can be desirable to make synonymous codon substitutions to match the codon usage pattern from the original organism in the heterologous expression host. Here we present CHARMING (for Codon HARMonizING), a robust and versatile algorithm to design mRNA sequences for heterologous gene expression and other related codon harmonization tasks. CHARMING can be run as a downloadable Python script or via a web portal at http://www.codons.org.

Supervised Prediction of Aging-Related Genes From a Context-Specific Protein Interaction Subnetwork.

Human aging is linked to many prevalent diseases. The aging process is highly influenced by genetic factors. Hence, it is important to identify human aging-related genes. We focus on supervised prediction of such genes. Gene expression-based methods for this purpose study genes in isolation from each other. While protein-protein interaction (PPI) network-based methods for this purpose account for interactions between genes' protein products, current PPI network data are context-unspecific, spanning different biological conditions. Instead, here, we focus on an aging-specific subnetwork of the entire PPI network, obtained by integrating aging-specific gene expression data and PPI network data. The potential of such data integration has been recognized but mostly in the context of cancer. So, we are the first to propose a supervised learning framework for predicting aging-related genes from an aging-specific PPI subnetwork. In a systematic and comprehensive evaluation, we find that in many of the evaluation tests: (i) using an aging-specific subnetwork indeed yields more accurate aging-related gene predictions than using the entire network, and (ii) predictive methods from our framework that have not previously been used for supervised prediction of aging-related genes outperform existing prominent methods for the same purpose. These results justify the need for our framework.

Towards future directions in data-integrative supervised prediction of human aging-related genes.

Motivation: Identification of human genes involved in the aging process is critical due to the incidence of many diseases with age. A state-of-the-art approach for this purpose infers a weighted dynamic aging-specific subnetwork by mapping gene expression (GE) levels at different ages onto the protein-protein interaction network (PPIN). Then, it analyzes this subnetwork in a supervised manner by training a predictive model to learn how network topologies of known aging- versus non-aging-related genes change across ages. Finally, it uses the trained model to predict novel aging-related gene candidates. However, the best current subnetwork resulting from this approach still yields suboptimal prediction accuracy. This could be because it was inferred using outdated GE and PPIN data. Here, we evaluate whether analyzing a weighted dynamic aging-specific subnetwork inferred from newer GE and PPIN data improves prediction accuracy upon analyzing the best current subnetwork inferred from outdated data. Results: Unexpectedly, we find that not to be the case. To understand this, we perform aging-related pathway and Gene Ontology term enrichment analyses. We find that the suboptimal prediction accuracy, regardless of which GE or PPIN data is used, may be caused by the current knowledge about which genes are aging-related being incomplete, or by the current methods for inferring or analyzing an aging-specific subnetwork being unable to capture all of the aging-related knowledge. These findings can potentially guide future directions towards improving supervised prediction of aging-related genes via -omics data integration. Availability and implementation: All data and code are available at zenodo, DOI: 10.5281/zenodo.6995045. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

Improved supervised prediction of aging-related genes via weighted dynamic network analysis.

BACKGROUND: This study focuses on the task of supervised prediction of aging-related genes from -omics data. Unlike gene expression methods for this task that capture aging-specific information but ignore interactions between genes (i.e., their protein products), or protein-protein interaction (PPI) network methods for this task that account for PPIs but the PPIs are context-unspecific, we recently integrated the two data types into an aging-specific PPI subnetwork, which yielded more accurate aging-related gene predictions. However, a dynamic aging-specific subnetwork did not improve prediction performance compared to a static aging-specific subnetwork, despite the aging process being dynamic. This could be because the dynamic subnetwork was inferred using a naive Induced subgraph approach. Instead, we recently inferred a dynamic aging-specific subnetwork using a methodologically more advanced notion of network propagation (NP), which improved upon Induced dynamic aging-specific subnetwork in a different task, that of unsupervised analyses of the aging process. RESULTS: Here, we evaluate whether our existing NP-based dynamic subnetwork will improve upon the dynamic as well as static subnetwork constructed by the Induced approach in the considered task of supervised prediction of aging-related genes. The existing NP-based subnetwork is unweighted, i.e., it gives equal importance to each of the aging-specific PPIs. Because accounting for aging-specific edge weights might be important, we additionally propose a weighted NP-based dynamic aging-specific subnetwork. We demonstrate that a predictive machine learning model trained and tested on the weighted subnetwork yields higher accuracy when predicting aging-related genes than predictive models run on the existing unweighted dynamic or static subnetworks, regardless of whether the existing subnetworks were inferred using NP or the Induced approach. CONCLUSIONS: Our proposed weighted dynamic aging-specific subnetwork and its corresponding predictive model could guide with higher confidence than the existing data and models the discovery of novel aging-related gene candidates for future wet lab validation.

Data-driven biological network alignment that uses topological, sequence, and functional information.

BACKGROUND: Network alignment (NA) can transfer functional knowledge between species' conserved biological network regions. Traditional NA assumes that it is topological similarity (isomorphic-like matching) between network regions that corresponds to the regions' functional relatedness. However, we recently found that functionally unrelated proteins are as topologically similar as functionally related proteins. So, we redefined NA as a data-driven method called TARA, which learns from network and protein functional data what kind of topological relatedness (rather than similarity) between proteins corresponds to their functional relatedness. TARA used topological information (within each network) but not sequence information (between proteins across networks). Yet, TARA yielded higher protein functional prediction accuracy than existing NA methods, even those that used both topological and sequence information. RESULTS: Here, we propose TARA++ that is also data-driven, like TARA and unlike other existing methods, but that uses across-network sequence information on top of within-network topological information, unlike TARA. To deal with the within-and-across-network analysis, we adapt social network embedding to the problem of biological NA. TARA++ outperforms protein functional prediction accuracy of existing methods. CONCLUSIONS: As such, combining research knowledge from different domains is promising. Overall, improvements in protein functional prediction have biomedical implications, for example allowing researchers to better understand how cancer progresses or how humans age.

Network-based protein structural classification.

Experimental determination of protein function is resource-consuming. As an alternative, computational prediction of protein function has received attention. In this context, protein structural classification (PSC) can help, by allowing for determining structural classes of currently unclassified proteins based on their features, and then relying on the fact that proteins with similar structures have similar functions. Existing PSC approaches rely on sequence-based or direct three-dimensional (3D) structure-based protein features. By contrast, we first model 3D structures of proteins as protein structure networks (PSNs). Then, we use network-based features for PSC. We propose the use of graphlets, state-of-the-art features in many research areas of network science, in the task of PSC. Moreover, because graphlets can deal only with unweighted PSNs, and because accounting for edge weights when constructing PSNs could improve PSC accuracy, we also propose a deep learning framework that automatically learns network features from weighted PSNs. When evaluated on a large set of approximately 9400 CATH and approximately 12 800 SCOP protein domains (spanning 36 PSN sets), the best of our proposed approaches are superior to existing PSC approaches in terms of accuracy, with comparable running times. Our data and code are available at https://doi.org/10.5281/zenodo.3787922.

Author Correction: GRAFENE: Graphlet-based alignment-free network approach integrates 3D structural and sequence (residue order) data to improve protein structural comparison.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Data-driven network alignment.

In this study, we deal with the problem of biological network alignment (NA), which aims to find a node mapping between species' molecular networks that uncovers similar network regions, thus allowing for the transfer of functional knowledge between the aligned nodes. We provide evidence that current NA methods, which assume that topologically similar nodes (i.e., nodes whose network neighborhoods are isomorphic-like) have high functional relatedness, do not actually end up aligning functionally related nodes. That is, we show that the current topological similarity assumption does not hold well. Consequently, we argue that a paradigm shift is needed with how the NA problem is approached. So, we redefine NA as a data-driven framework, called TARA (data-driven NA), which attempts to learn the relationship between topological relatedness and functional relatedness without assuming that topological relatedness corresponds to topological similarity. TARA makes no assumptions about what nodes should be aligned, distinguishing it from existing NA methods. Specifically, TARA trains a classifier to predict whether two nodes from different networks are functionally related based on their network topological patterns (features). We find that TARA is able to make accurate predictions. TARA then takes each pair of nodes that are predicted as related to be part of an alignment. Like traditional NA methods, TARA uses this alignment for the across-species transfer of functional knowledge. TARA as currently implemented uses topological but not protein sequence information for functional knowledge transfer. In this context, we find that TARA outperforms existing state-of-the-art NA methods that also use topological information, WAVE and SANA, and even outperforms or complements a state-of-the-art NA method that uses both topological and sequence information, PrimAlign. Hence, adding sequence information to TARA, which is our future work, is likely to further improve its performance. The software and data are available at http://www.nd.edu/~cone/TARA/.

Network analysis of synonymous codon usage.

MOTIVATION: Most amino acids are encoded by multiple synonymous codons, some of which are used more rarely than others. Analyses of positions of such rare codons in protein sequences revealed that rare codons can impact co-translational protein folding and that positions of some rare codons are evolutionarily conserved. Analyses of their positions in protein 3-dimensional structures, which are richer in biochemical information than sequences alone, might further explain the role of rare codons in protein folding. RESULTS: We model protein structures as networks and use network centrality to measure the structural position of an amino acid. We first validate that amino acids buried within the structural core are network-central, and those on the surface are not. Then, we study potential differences between network centralities and thus structural positions of amino acids encoded by conserved rare, non-conserved rare and commonly used codons. We find that in 84% of proteins, the three codon categories occupy significantly different structural positions. We examine protein groups showing different codon centrality trends, i.e. different relationships between structural positions of the three codon categories. We see several cases of all proteins from our data with some structural or functional property being in the same group. Also, we see a case of all proteins in some group having the same property. Our work shows that codon usage is linked to the final protein structure and thus possibly to co-translational protein folding. AVAILABILITY AND IMPLEMENTATION: https://nd.edu/∼cone/CodonUsage/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Analysis of computational codon usage models and their association with translationally slow codons.

Improved computational modeling of protein translation rates, including better prediction of where translational slowdowns along an mRNA sequence may occur, is critical for understanding co-translational folding. Because codons within a synonymous codon group are translated at different rates, many computational translation models rely on analyzing synonymous codons. Some models rely on genome-wide codon usage bias (CUB), believing that globally rare and common codons are the most informative of slow and fast translation, respectively. Others use the CUB observed only in highly expressed genes, which should be under selective pressure to be translated efficiently (and whose CUB may therefore be more indicative of translation rates). No prior work has analyzed these models for their ability to predict translational slowdowns. Here, we evaluate five models for their association with slowly translated positions as denoted by two independent ribosome footprint (RFP) count experiments from S. cerevisiae, because RFP data is often considered as a "ground truth" for translation rates across mRNA sequences. We show that all five considered models strongly associate with the RFP data and therefore have potential for estimating translational slowdowns. However, we also show that there is a weak correlation between RFP counts for the same genes originating from independent experiments, even when their experimental conditions are similar. This raises concerns about the efficacy of using current RFP experimental data for estimating translation rates and highlights a potential advantage of using computational models to understand translation rates instead.

L-HetNetAligner: A novel algorithm for Local Alignment of Heterogeneous Biological Networks.

Networks are largely used for modelling and analysing a wide range of biological data. As a consequence, many different research efforts have resulted in the introduction of a large number of algorithms for analysis and comparison of networks. Many of these algorithms can deal with networks with a single class of nodes and edges, also referred to as homogeneous networks. Recently, many different approaches tried to integrate into a single model the interplay of different molecules. A possible formalism to model such a scenario comes from node/edge coloured networks (also known as heterogeneous networks) implemented as node/ edge-coloured graphs. Therefore, the need for the introduction of algorithms able to compare heterogeneous networks arises. We here focus on the local comparison of heterogeneous networks, and we formulate it as a network alignment problem. To the best of our knowledge, the local alignment of heterogeneous networks has not been explored in the past. We here propose L-HetNetAligner a novel algorithm that receives as input two heterogeneous networks (node-coloured graphs) and builds a local alignment of them. We also implemented and tested our algorithm. Our results confirm that our method builds high-quality alignments. The following website *contains Supplementary File 1 material and the code.

Pairwise Versus Multiple Global Network Alignment.

Biological network alignment (NA) aims to identify similar regions between molecular networks of different species. NA can be local or global. Just as the recent trend in the NA field, we also focus on global NA, which can be pairwise (PNA) and multiple (MNA). PNA produces aligned node pairs between two networks. MNA produces aligned node clusters between more than two networks. Recently, the focus has shifted from PNA to MNA, because MNA captures conserved regions between more networks than PNA (and MNA is thus hypothesized to yield higher-quality alignments), though at higher computational complexity. The issue is that, due to the different outputs of PNA and MNA, a PNA method is only compared to other PNA methods, and an MNA method is only compared to other MNA methods. Comparison of PNA against MNA must be done to evaluate whether MNA indeed yields higher-quality alignments, as only this would justify MNA's higher computational complexity. We introduce a framework that allows for this. We evaluate eight prominent PNA and MNA methods, on synthetic and real-world biological networks, using topological and functional alignment quality measures. We compare PNA against MNA in both a pairwise (native to PNA) and multiple (native to MNA) manner. PNA is expected to perform better under the pairwise evaluation framework. Indeed this is what we find. MNA is expected to perform better under the multiple evaluation framework. Shockingly, we find this not always to hold; PNA is often better than MNA in this framework, depending on the choice of evaluation test.

The power of dynamic social networks to predict individuals' mental health.

Precision medicine has received attention both in and outside the clinic. We focus on the latter, by exploiting the relationship between individuals' social interactions and their mental health to predict one's likelihood of being depressed or anxious from rich dynamic social network data. Existing studies differ from our work in at least one aspect: they do not model social interaction data as a network; they do so but analyze static network data; they examine "correlation" between social networks and health but without making any predictions; or they study other individual traits but not mental health. In a comprehensive evaluation, we show that our predictive model that uses dynamic social network data is superior to its static network as well as non-network equivalents when run on the same data. Supplementary material for this work is available at https://nd.edu/~cone/NetHealth/PSB_SM.pdf.

Temporal network alignment via GoT-WAVE.

MOTIVATION: Network alignment (NA) finds conserved regions between two networks. NA methods optimize node conservation (NC) and edge conservation. Dynamic graphlet degree vectors are a state-of-the-art dynamic NC measure, used within the fastest and most accurate NA method for temporal networks: DynaWAVE. Here, we use graphlet-orbit transitions (GoTs), a different graphlet-based measure of temporal node similarity, as a new dynamic NC measure within DynaWAVE, resulting in GoT-WAVE. RESULTS: On synthetic networks, GoT-WAVE improves DynaWAVE's accuracy by 30% and speed by 64%. On real networks, when optimizing only dynamic NC, the methods are complementary. Furthermore, only GoT-WAVE supports directed edges. Hence, GoT-WAVE is a promising new temporal NA algorithm, which efficiently optimizes dynamic NC. We provide a user-friendly user interface and source code for GoT-WAVE. AVAILABILITY AND IMPLEMENTATION: http://www.dcc.fc.up.pt/got-wave/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Network analysis of the NetHealth data: exploring co-evolution of individuals' social network positions and physical activities.

Understanding the relationship between individuals' social networks and health could help devise public health interventions for reducing incidence of unhealthy behaviors or increasing prevalence of healthy ones. In this context, we explore the co-evolution of individuals' social network positions and physical activities. We are able to do so because the NetHealth study at the University of Notre Dame has generated both high-resolution longitudinal social network (e.g., SMS) data and high-resolution longitudinal health-related behavioral (e.g., Fitbit physical activity) data. We examine trait differences between (i) users whose social network positions (i.e., centralities) change over time versus those whose centralities remain stable, (ii) users whose Fitbit physical activities change over time versus those whose physical activities remain stable, and (iii) users whose centralities and their physical activities co-evolve, i.e., correlate with each other over time. We find that centralities of a majority of all nodes change with time. These users do not show any trait difference compared to time-stable users. However, if out of all users whose centralities change with time we focus on those whose physical activities also change with time, then the resulting users are more likely to be introverted than time-stable users. Moreover, users whose centralities and physical activities both change with time and whose evolving centralities are significantly correlated (i.e., co-evolve) with evolving physical activities are more likely to be introverted as well as anxious compared to those users who are time-stable and do not have a co-evolution relationship. Our network analysis framework reveals several links between individuals' social network structure, health-related behaviors, and the other (e.g., personality) traits. In the future, our study could lead to development of a predictive model of social network structure from behavioral/trait information and vice versa.

From homogeneous to heterogeneous network alignment via colored graphlets.

Network alignment (NA) compares networks with the goal of finding a node mapping that uncovers highly similar (conserved) network regions. Existing NA methods are homogeneous, i.e., they can deal only with networks containing nodes and edges of one type. Due to increasing amounts of heterogeneous network data with nodes or edges of different types, we extend three recent state-of-the-art homogeneous NA methods, WAVE, MAGNA++, and SANA, to allow for heterogeneous NA for the first time. We introduce several algorithmic novelties. Namely, these existing methods compute homogeneous graphlet-based node similarities and then find high-scoring alignments with respect to these similarities, while simultaneously maximizing the amount of conserved edges. Instead, we extend homogeneous graphlets to their heterogeneous counterparts, which we then use to develop a new measure of heterogeneous node similarity. Also, we extend S3, a state-of-the-art measure of edge conservation for homogeneous NA, to its heterogeneous counterpart. Then, we find high-scoring alignments with respect to our heterogeneous node similarity and edge conservation measures. In evaluations on synthetic and real-world biological networks, our proposed heterogeneous NA methods lead to higher-quality alignments and better robustness to noise in the data than their homogeneous counterparts. The software and data from this work is available at https://nd.edu/~cone/colored_graphlets/.

ClueNet: Clustering a temporal network based on topological similarity rather than denseness.

Network clustering is a very popular topic in the network science field. Its goal is to divide (partition) the network into groups (clusters or communities) of "topologically related" nodes, where the resulting topology-based clusters are expected to "correlate" well with node label information, i.e., metadata, such as cellular functions of genes/proteins in biological networks, or age or gender of people in social networks. Even for static data, the problem of network clustering is complex. For dynamic data, the problem is even more complex, due to an additional dimension of the data-their temporal (evolving) nature. Since the problem is computationally intractable, heuristic approaches need to be sought. Existing approaches for dynamic network clustering (DNC) have drawbacks. First, they assume that nodes should be in the same cluster if they are densely interconnected within the network. We hypothesize that in some applications, it might be of interest to cluster nodes that are topologically similar to each other instead of or in addition to requiring the nodes to be densely interconnected. Second, they ignore temporal information in their early steps, and when they do consider this information later on, they do so implicitly. We hypothesize that capturing temporal information earlier in the clustering process and doing so explicitly will improve results. We test these two hypotheses via our new approach called ClueNet. We evaluate ClueNet against six existing DNC methods on both social networks capturing evolving interactions between individuals (such as interactions between students in a high school) and biological networks capturing interactions between biomolecules in the cell at different ages. We find that ClueNet is superior in over 83% of all evaluation tests. As more real-world dynamic data are becoming available, DNC and thus ClueNet will only continue to gain importance.