Dynamic modularity in protein interaction networks predicts breast cancer outcome

Community cohesion looseness in gene networks reveals individualized drug targets and resistance

Community cohesion plays a critical role in the determination of an individual's health in social science. Intriguingly, a community structure of gene networks indicates that the concept of community cohesion could be applied between the genes as well to overcome the limitations of single gene-based biomarkers for precision oncology. Here, we develop community cohesion scores which precisely quantify the community ability to retain the interactions between the genes and their cellular functions in each individualized gene network. Using breast cancer as a proof-of-concept study, we measure the community cohesion score profiles of 950 case samples and predict the individualized therapeutic targets in 2-fold. First, we prioritize them by finding druggable genes present in the community with the most and relatively decreased scores in each individual. Then, we pinpoint more individualized therapeutic targets by discovering the genes which greatly contribute to the community cohesion looseness in each individualized gene network. Compared with the previous approaches, the community cohesion scores show at least four times higher performance in predicting effective individualized chemotherapy targets based on drug sensitivity data. Furthermore, the community cohesion scores successfully discover the known breast cancer subtypes and we suggest new targeted therapy targets for triple negative breast cancer (e.g. KIT and GABRP). Lastly, we demonstrate that the community cohesion scores can predict tamoxifen responses in ER+ breast cancer and suggest potential combination therapies (e.g. NAMPT and RXRA inhibitors) to reduce endocrine therapy resistance based on individualized characteristics. Our method opens new perspectives for the biomarker development in precision oncology.

Enabling personalised disease diagnosis by combining a patient's time-specific gene expression profile with a biomedical knowledge base

BACKGROUND

Recent developments in the domain of biomedical knowledge bases (KBs) open up new ways to exploit biomedical knowledge that is available in the form of KBs. Significant work has been done in the direction of biomedical KB creation and KB completion, specifically, those having gene-disease associations and other related entities. However, the use of such biomedical KBs in combination with patients' temporal clinical data still largely remains unexplored, but has the potential to immensely benefit medical diagnostic decision support systems.

RESULTS

We propose two new algorithms, LOADDx and SCADDx, to combine a patient's gene expression data with gene-disease association and other related information available in the form of a KB, to assist personalized disease diagnosis. We have tested both of the algorithms on two KBs and on four real-world gene expression datasets of respiratory viral infection caused by Influenza-like viruses of 19 subtypes. We also compare the performance of proposed algorithms with that of five existing state-of-the-art machine learning algorithms (k-NN, Random Forest, XGBoost, Linear SVM, and SVM with RBF Kernel) using two validation approaches: LOOCV and a single internal validation set. Both SCADDx and LOADDx outperform the existing algorithms when evaluated with both validation approaches. SCADDx is able to detect infections with up to 100% accuracy in the cases of Datasets 2 and 3. Overall, SCADDx and LOADDx are able to detect an infection within 72 h of infection with 91.38% and 92.66% average accuracy respectively considering all four datasets, whereas XGBoost, which performed best among the existing machine learning algorithms, can detect the infection with only 86.43% accuracy on an average.

CONCLUSIONS

We demonstrate how our novel idea of using the most and least differentially expressed genes in combination with a KB can enable identification of the diseases that a patient is most likely to have at a particular time, from a KB with thousands of diseases. Moreover, the proposed algorithms can provide a short ranked list of the most likely diseases for each patient along with their most affected genes, and other entities linked with them in the KB, which can support health care professionals in their decision-making.

Hubs and Bottlenecks in Protein-Protein Interaction Networks

Protein-protein interaction networks (PPINs) represent the physical interactions among proteins in a cell. These interactions are critical in all cellular processes, including signal transduction, metabolic regulation, and gene expression. In PPINs, centrality measures are widely used to identify the most critical nodes. The two most commonly used centrality measures in networks are degree and betweenness centralities. Degree centrality is the number of connections a node has in the network, and betweenness centrality is the measure of the extent to which a node lies on the shortest paths between pairs of other nodes in the network. In PPINs, proteins with high degree and betweenness centrality are referred to as hubs and bottlenecks respectively. Hubs and bottlenecks are topologically and functionally essential proteins that play crucial roles in maintaining the network's structure and function. This article comprehensively reviews essential literature on hubs and bottlenecks, including their properties and functions.

Protein-protein interaction network module changes associated with the vertebrate fin-to-limb transition

Evolutionary phenotypic transitions, such as the fin-to-limb transition in vertebrates, result from modifications in related proteins and their interactions, often in response to changing environment. Identifying these alterations in protein networks is crucial for a more comprehensive understanding of these transitions. However, previous research has not attempted to compare protein-protein interaction (PPI) networks associated with evolutionary transitions, and most experimental studies concentrate on a limited set of proteins. Therefore, the goal of this work was to develop a network-based platform for investigating the fin-to-limb transition using PPI networks. Quality-enhanced protein networks, constructed by integrating PPI networks with anatomy ontology data, were leveraged to compare protein modules for paired fins (pectoral fin and pelvic fin) of fishes (zebrafish) to those of the paired limbs (forelimb and hindlimb) of mammals (mouse). This also included prediction of novel protein candidates and their validation by enrichment and homology analyses. Hub proteins such as shh and bmp4, which are crucial for module stability, were identified, and their changing roles throughout the transition were examined. Proteins with preserved roles during the fin-to-limb transition were more likely to be hub proteins. This study also addressed hypotheses regarding the role of non-preserved proteins associated with the transition.

Identifying Lymph Node Metastasis-Related Factors in Breast Cancer Using Differential Modular and Mutational Structural Analysis

Complex diseases are generally caused by disorders of biological networks and/or mutations in multiple genes. Comparisons of network topologies between different disease states can highlight key factors in their dynamic processes. Here, we propose a differential modular analysis approach that integrates protein-protein interactions with gene expression profiles for modular analysis, and introduces inter-modular edges and date hubs to identify the "core network module" that quantifies the significant phenotypic variation. Then, based on this core network module, key factors, including functional protein-protein interactions, pathways, and driver mutations, are predicted by the topological-functional connection score and structural modeling. We applied this approach to analyze the lymph node metastasis (LNM) process in breast cancer. The functional enrichment analysis showed that both inter-modular edges and date hubs play important roles in cancer metastasis and invasion, and in metastasis hallmarks. The structural mutation analysis suggested that the LNM of breast cancer may be the outcome of the dysfunction of rearranged during transfection (RET) proto-oncogene-related interactions and the non-canonical calcium signaling pathway via an allosteric mutation of RET. We believe that the proposed method can provide new insights into disease progression such as cancer metastasis.

Robust, scalable, and informative clustering for diverse biological networks

Clustering molecular data into informative groups is a primary step in extracting robust conclusions from big data. However, due to foundational issues in how they are defined and detected, such clusters are not always reliable, leading to unstable conclusions. We compare popular clustering algorithms across thousands of synthetic and real biological datasets, including a new consensus clustering algorithm-SpeakEasy2: Champagne. These tests identify trends in performance, show no single method is universally optimal, and allow us to examine factors behind variation in performance. Multiple metrics indicate SpeakEasy2 generally provides robust, scalable, and informative clusters for a range of applications.

Using a stacked ensemble learning framework to predict modulators of protein-protein interactions

Identifying small molecule protein-protein interaction modulators (PPIMs) is a highly promising and meaningful research direction for drug discovery, cancer treatment, and other fields. In this study, we developed a stacking ensemble computational framework, SELPPI, based on a genetic algorithm and tree-based machine learning method for effectively predicting new modulators targeting protein-protein interactions. More specifically, extremely randomized trees (ExtraTrees), adaptive boosting (AdaBoost), random forest (RF), cascade forest, light gradient boosting machine (LightGBM), and extreme gradient boosting (XGBoost) were used as basic learners. Seven types of chemical descriptors were taken as the input characteristic parameters. Primary predictions were obtained with each basic learner-descriptor pair. Then, the 6 methods mentioned above were used as meta learners and trained on the primary prediction in turn. The most efficient method was utilized as the meta learner. Finally, the genetic algorithm was used to select the optimal primary prediction output as the input of the meta learner for secondary prediction to obtain the final result. We systematically evaluated our model on the pdCSM-PPI datasets. To our knowledge, our model outperformed all existing models, which demonstrates its great power.

Dissection of hubs and bottlenecks in a protein-protein interaction network

Analysis of degree centrality in conjunction with betweenness centrality of proteins in a human protein-protein interaction network revealed three categories of centrally important proteins: a) proteins with high degree and betweenness (hub-bottlenecks denoted as MX), b) proteins with high betweenness and low degree (non-hub-bottlenecks/pure bottlenecks denoted as PB) and c) proteins with high degree and low betweenness (hub-non-bottlenecks/pure hubs denoted as PH). When subjected to a detailed statistical analysis of their molecular-level properties, the proteins belonging to each of these categories were found to be associated with distinct canonical molecular properties, i.e., "molecular markers". The MX proteins are a) conformationally versatile, mainly comprising of essential proteins, b) the targets for interactions by the proteins of viral and bacterial pathogens, c) evolutionally constrained, involved in multiple pathways, enriched with disease genes and d) involved in the functions such as protein stabilization, phosphorylation, and mRNA slicing processes. PB proteins are a) enriched with extracellular and cancer-related proteins, b) enriched with the approved drug targets and c) involved in cell-cell signaling processes. Finally, PH are a) structurally versatile, b) enriched with essential proteins primarily involved in housekeeping processes (transcription and replication). The fact that the proteins belonging to these three categories form three distinct sets in terms of their molecular properties reveals the existence of trichotomy among hubs and bottlenecks, and this knowledge is of paramount importance while prioritizing protein targets for further studies such as drug design and disease association studies based on their network centrality values.

Discovering Entities Similarities in Biological Networks Using a Hybrid Immune Algorithm

Disease phenotypes are generally caused by the failure of gene modules which often have similar biological roles. Through the study of biological networks, it is possible to identify the intrinsic structure of molecular interactions in order to identify the so-called “disease modules”. Community detection is an interesting and valuable approach to discovering the structure of the community in a complex network, revealing the internal organization of the nodes, and has become a leading research topic in the analysis of complex networks. This work investigates the link between biological modules and network communities in test-case biological networks that are commonly used as a reference point and which include Protein–Protein Interaction Networks, Metabolic Networks and Transcriptional Regulation Networks. In order to identify small and structurally well-defined communities in the biological context, a hybrid immune metaheuristic algorithm Hybrid-IA is proposed and compared with several metaheuristics, hyper-heuristics, and the well-known greedy algorithm Louvain, with respect to modularity maximization. Considering the limitation of modularity optimization, which can fail to identify smaller communities, the reliability of Hybrid-IA was also analyzed with respect to three well-known sensitivity analysis measures (NMI, ARI and NVI) that assess how similar the detected communities are to real ones. By inspecting all outcomes and the performed comparisons, we will see that on one hand Hybrid-IA finds slightly lower modularity values than Louvain, but outperforms all other metaheuristics, while on the other hand, it can detect communities more similar to the real ones when compared to those detected by Louvain.

Partial synchronization and community switching in phase-oscillator networks and its analysis based on a bidirectional, weighted chain of three oscillators

Complex networks often possess communities defined based on network connectivity. When dynamics undergo in a network, one can also consider dynamical communities, i.e., a group of nodes displaying a similar dynamical process. We have investigated both analytically and numerically the development of a dynamical community structure, where the community is referred to as a group of nodes synchronized in frequency, in networks of phase oscillators. We first demonstrate that using a few example networks, the community structure changes when network connectivity or interaction strength is varied. In particular, we found that community switching, i.e., a portion of oscillators change the group to which they synchronize, occurs for a range of parameters. We then propose a three-oscillator model: a bidirectional, weighted chain of three Kuramoto phase oscillators, as a theoretical framework for understanding the community formation and its variation. Our analysis demonstrates that the model shows a variety of partially synchronized patterns: oscillators with similar natural frequencies tend to synchronize for weak coupling, while tightly connected oscillators tend to synchronize for strong coupling. We obtain approximate expressions for the critical coupling strengths by employing a perturbative approach in a weak coupling regime and a geometric approach in strong coupling regimes. Moreover, we elucidate the bifurcation types of transitions between different patterns. Our theory might be useful for understanding the development of partially synchronized patterns in a wider class of complex networks than community structured networks.

Inferencing Bulk Tumor and Single-Cell Multi-Omics Regulatory Networks for Discovery of Biomarkers and Therapeutic Targets

There are insufficient accurate biomarkers and effective therapeutic targets in current cancer treatment. Multi-omics regulatory networks in patient bulk tumors and single cells can shed light on molecular disease mechanisms. Integration of multi-omics data with large-scale patient electronic medical records (EMRs) can lead to the discovery of biomarkers and therapeutic targets. In this review, multi-omics data harmonization methods were introduced, and common approaches to molecular network inference were summarized. Our Prediction Logic Boolean Implication Networks (PLBINs) have advantages over other methods in constructing genome-scale multi-omics networks in bulk tumors and single cells in terms of computational efficiency, scalability, and accuracy. Based on the constructed multi-modal regulatory networks, graph theory network centrality metrics can be used in the prioritization of candidates for discovering biomarkers and therapeutic targets. Our approach to integrating multi-omics profiles in a patient cohort with large-scale patient EMRs such as the SEER-Medicare cancer registry combined with extensive external validation can identify potential biomarkers applicable in large patient populations. These methodologies form a conceptually innovative framework to analyze various available information from research laboratories and healthcare systems, accelerating the discovery of biomarkers and therapeutic targets to ultimately improve cancer patient survival outcomes.

Recent advances in the field of single-cell proteomics

The field of single-cell omics is rapidly progressing. Although DNA and RNA sequencing-based methods have dominated the field to date, global proteome profiling has also entered the main stage. Single-cell proteomics was facilitated by advancements in different aspects of mass spectrometry (MS)-based proteomics, such as instrument design, sample preparation, chromatography and ion mobility. Single-cell proteomics by mass spectrometry (scp-MS) has moved beyond being a mere technical development, and is now able to deliver actual biological application and has been successfully applied to characterize different cell states. Here, we review some key developments of scp-MS, provide a background to the field, discuss the various available methods and foresee possible future directions.

Computational systems biology in disease modeling and control, review and perspectives

Omics-based approaches have become increasingly influential in identifying disease mechanisms and drug responses. Considering that diseases and drug responses are co-expressed and regulated in the relevant omics data interactions, the traditional way of grabbing omics data from single isolated layers cannot always obtain valuable inference. Also, drugs have adverse effects that may impair patients, and launching new medicines for diseases is costly. To resolve the above difficulties, systems biology is applied to predict potential molecular interactions by integrating omics data from genomic, proteomic, transcriptional, and metabolic layers. Combined with known drug reactions, the resulting models improve medicines' therapeutical performance by re-purposing the existing drugs and combining drug molecules without off-target effects. Based on the identified computational models, drug administration control laws are designed to balance toxicity and efficacy. This review introduces biomedical applications and analyses of interactions among gene, protein and drug molecules for modeling disease mechanisms and drug responses. The therapeutical performance can be improved by combining the predictive and computational models with drug administration designed by control laws. The challenges are also discussed for its clinical uses in this work.

Large-scale prediction of key dynamic interacting proteins in multiple cancers

Tracking cancer dynamic protein-protein interactions (PPIs) and deciphering their pathogenesis remain a challenge. We presented a dynamic PPIs' hypothesis: permanent and transient interactions might achieve dynamic switchings from normal cells to malignancy, which could cause maintenance functions to be interrupted and transient functions to be sustained. Based on the hypothesis, we first predicted >1400 key cancer genes (KCG) by applying PPI-express we proposed to 18 cancer gene expression datasets. We then further screened out key dynamic interactions (KDI) of cancer based on KCG and transient and permanent interactions under both conditions. Two prominent functional characteristics, "Cell cycle-related" and "Immune-related", were presented for KCG, suggesting that these might be their general characteristics. We found that, compared to permanent to transient KDI pairs (P2T) in the network, transient to permanent (T2P) have significantly higher edge betweenness (EB), and P2T pairs tending to locate intra-functional modules may play roles in maintaining normal biological functions, while T2P KDI pairs tending to locate inter-modules may play roles in biological signal transduction. It was consistent with our hypothesis. Also, we analyzed network characteristics of KDI pairs and their functions. Our findings of KDI may serve to understand and explain a few hallmarks of cancer.

From the origin of life to pandemics: emergent phenomena in complex systems

When a large number of similar entities interact among each other and with their environment at a low scale, unexpected outcomes at higher spatio-temporal scales might spontaneously arise. This non-trivial phenomenon, known as emergence, characterizes a broad range of distinct complex systems-from physical to biological and social-and is often related to collective behaviour. It is ubiquitous, from non-living entities such as oscillators that under specific conditions synchronize, to living ones, such as birds flocking or fish schooling. Despite the ample phenomenological evidence of the existence of systems' emergent properties, central theoretical questions to the study of emergence remain unanswered, such as the lack of a widely accepted, rigorous definition of the phenomenon or the identification of the essential physical conditions that favour emergence. We offer here a general overview of the phenomenon of emergence and sketch current and future challenges on the topic. Our short review also serves as an introduction to the theme issue Emergent phenomena in complex physical and socio-technical systems: from cells to societies, where we provide a synthesis of the contents tackled in the issue and outline how they relate to these challenges, spanning from current advances in our understanding on the origin of life to the large-scale propagation of infectious diseases. This article is part of the theme issue 'Emergent phenomena in complex physical and socio-technical systems: from cells to societies'.

Artificial intelligence in cancer target identification and drug discovery

Artificial intelligence is an advanced method to identify novel anticancer targets and discover novel drugs from biology networks because the networks can effectively preserve and quantify the interaction between components of cell systems underlying human diseases such as cancer. Here, we review and discuss how to employ artificial intelligence approaches to identify novel anticancer targets and discover drugs. First, we describe the scope of artificial intelligence biology analysis for novel anticancer target investigations. Second, we review and discuss the basic principles and theory of commonly used network-based and machine learning-based artificial intelligence algorithms. Finally, we showcase the applications of artificial intelligence approaches in cancer target identification and drug discovery. Taken together, the artificial intelligence models have provided us with a quantitative framework to study the relationship between network characteristics and cancer, thereby leading to the identification of potential anticancer targets and the discovery of novel drug candidates.

A network-based approach to identify expression modules underlying rejection in pediatric liver transplantation

Selecting the right immunosuppressant to ensure rejection-free outcomes poses unique challenges in pediatric liver transplant (LT) recipients. A molecular predictor can comprehensively address these challenges. Currently, there are no well-validated blood-based biomarkers for pediatric LT recipients before or after LT. Here, we discover and validate separate pre- and post-LT transcriptomic signatures of rejection. Using an integrative machine learning approach, we combine transcriptomics data with the reference high-quality human protein interactome to identify network module signatures, which underlie rejection. Unlike gene signatures, our approach is inherently multivariate and more robust to replication and captures the structure of the underlying network, encapsulating additive effects. We also identify, in an individual-specific manner, signatures that can be targeted by current anti-rejection drugs and other drugs that can be repurposed. Our approach can enable personalized adjustment of drug regimens for the dominant targetable pathways before and after LT in children.

Oscillation suppression and chimera states in time-varying networks

Complex network theory has offered a powerful platform for the study of several natural dynamic scenarios, based on the synergy between the interaction topology and the dynamics of its constituents. With research in network theory being developed so fast, it has become extremely necessary to move from simple network topologies to more sophisticated and realistic descriptions of the connectivity patterns. In this context, there is a significant amount of recent works that have emerged with enormous evidence establishing the time-varying nature of the connections among the constituents in a large number of physical, biological, and social systems. The recent review article by Ghosh et al. [Phys. Rep. 949, 1-63 (2022)] demonstrates the significance of the analysis of collective dynamics arising in temporal networks. Specifically, the authors put forward a detailed excerpt of results on the origin and stability of synchronization in time-varying networked systems. However, among the complex collective dynamical behaviors, the study of the phenomenon of oscillation suppression and that of other diverse aspects of synchronization are also considered to be central to our perception of the dynamical processes over networks. Through this review, we discuss the principal findings from the research studies dedicated to the exploration of the two collective states, namely, oscillation suppression and chimera on top of time-varying networks of both static and mobile nodes. We delineate how temporality in interactions can suppress oscillation and induce chimeric patterns in networked dynamical systems, from effective analytical approaches to computational aspects, which is described while addressing these two phenomena. We further sketch promising directions for future research on these emerging collective behaviors in time-varying networks.

Empowering differential networks using Bayesian analysis

Differential networks (DN) are important tools for modeling the changes in conditional dependencies between multiple samples. A Bayesian approach for estimating DNs, from the classical viewpoint, is introduced with a computationally efficient threshold selection for graphical model determination. The algorithm separately estimates the precision matrices of the DN using the Bayesian adaptive graphical lasso procedure. Synthetic experiments illustrate that the Bayesian DN performs exceptionally well in numerical accuracy and graphical structure determination in comparison to state of the art methods. The proposed method is applied to South African COVID-19 data to investigate the change in DN structure between various phases of the pandemic.

Single-Step Affinity Purification (ssAP) and Mass Spectrometry of Macromolecular Complexes in the Yeast S. cerevisiae

Cellular functions are mostly defined by the dynamic interactions of proteins within macromolecular networks. Deciphering the composition of macromolecular complexes and their dynamic rearrangements is the key to get a comprehensive picture of cellular behavior and to understand biological systems. In the past two decades, affinity purification coupled to mass spectrometry has become a powerful tool to comprehensively study interaction networks and their assemblies. To overcome initial limitations of the approach, in particular, the effect of protein and RNA degradation, loss of transient interactors, and poor overall yield of intact complexes from cell lysates, various modifications to affinity purification protocols have been devised over the years. In this chapter, we describe a rapid single-step affinity purification method for the efficient isolation of dynamic macromolecular complexes. The technique employs cell lysis by cryo-milling, which ensures nondegraded starting material in the submicron range, and magnetic beads, which allow for dense antibody-conjugation and thus rapid complex isolation, while avoiding loss of transient interactions. The method is epitope tag-independent, and overcomes many of the previous limitations to produce large interactomes with almost no contamination. The protocol as described here has been optimized for the yeast S. cerevisiae.

Are we there yet? A machine learning architecture to predict organotropic metastases

BACKGROUND & AIMS

Cancer metastasis into distant organs is an evolutionarily selective process. A better understanding of the driving forces endowing proliferative plasticity of tumor seeds in distant soils is required to develop and adapt better treatment systems for this lethal stage of the disease. To this end, we aimed to utilize transcript expression profiling features to predict the site-specific metastases of primary tumors and second, to identify the determinants of tissue specific progression.

METHODS

We used statistical machine learning for transcript feature selection to optimize classification and built tree-based classifiers to predict tissue specific sites of metastatic progression.

RESULTS

We developed a novel machine learning architecture that analyzes 33 types of RNA transcriptome profiles from The Cancer Genome Atlas (TCGA) database. Our classifier identifies the tumor type, derives synthetic instances of primary tumors metastasizing to distant organs and classifies the site-specific metastases in 16 types of cancers metastasizing to 12 locations.

CONCLUSIONS

We have demonstrated that site specific metastatic progression is predictable using transcriptomic profiling data from primary tumors and that the overrepresented biological processes in tumors metastasizing to congruent distant loci are highly overlapping. These results indicate site-specific progression was organotropic and core features of biological signaling pathways are identifiable that may describe proliferative plasticity in distant soils.

Computationally Reconstructed Interactome of Bradyrhizobium diazoefficiens USDA110 Reveals Novel Functional Modules and Protein Hubs for Symbiotic Nitrogen Fixation

Symbiotic nitrogen fixation is an important part of the nitrogen biogeochemical cycles and the main nitrogen source of the biosphere. As a classical model system for symbiotic nitrogen fixation, rhizobium-legume systems have been studied elaborately for decades. Details about the molecular mechanisms of the communication and coordination between rhizobia and host plants is becoming clearer. For more systematic insights, there is an increasing demand for new studies integrating multiomics information. Here, we present a comprehensive computational framework integrating the reconstructed protein interactome of B. diazoefficiens USDA110 with its transcriptome and proteome data to study the complex protein-protein interaction (PPI) network involved in the symbiosis system. We reconstructed the interactome of B. diazoefficiens USDA110 by computational approaches. Based on the comparison of interactomes between B. diazoefficiens USDA110 and other rhizobia, we inferred that the slow growth of B. diazoefficiens USDA110 may be due to the requirement of more protein modifications, and we further identified 36 conserved functional PPI modules. Integrated with transcriptome and proteome data, interactomes representing free-living cell and symbiotic nitrogen-fixing (SNF) bacteroid were obtained. Based on the SNF interactome, a core-sub-PPI-network for symbiotic nitrogen fixation was determined and nine novel functional modules and eleven key protein hubs playing key roles in symbiosis were identified. The reconstructed interactome of B. diazoefficiens USDA110 may serve as a valuable reference for studying the mechanism underlying the SNF system of rhizobia and legumes.

Analysis of multiple gene co-expression networks to discover interactions favoring CFTR biogenesis and ΔF508-CFTR rescue

BACKGROUND

We previously reported that expression of a miR-138 mimic or knockdown of SIN3A in primary cultures of cystic fibrosis (CF) airway epithelia increased ΔF508-CFTR mRNA and protein levels, and partially restored CFTR-dependent chloride transport. Global mRNA transcript profiling in ΔF508-CFBE cells treated with miR-138 mimic or SIN3A siRNA identified two genes, SYVN1 and NEDD8, whose inhibition significantly increased ΔF508-CFTR trafficking, maturation, and function. Little is known regarding the dynamic changes in the CFTR gene network during such rescue events. We hypothesized that analysis of condition-specific gene networks from transcriptomic data characterizing ΔF508-CFTR rescue could help identify dynamic gene modules associated with CFTR biogenesis.

METHODS

We applied a computational method, termed M-module, to analyze multiple gene networks, each of which exhibited differential activity compared to a baseline condition. In doing so, we identified both unique and shared gene pathways across multiple differential networks. To construct differential networks, gene expression data from CFBE cells were divided into three groups: (1) siRNA inhibition of NEDD8 and SYVN1; (2) miR-138 mimic and SIN3A siRNA; and (3) temperature (27 °C for 24 h, 40 °C for 24 h, and 27 °C for 24 h followed by 40 °C for 24 h).

RESULTS

Interrogation of individual networks (e.g., NEDD8/SYVN1 network), combinations of two networks (e.g., NEDD8/SYVN1 + temperature networks), and all three networks yielded sets of 1-modules, 2-modules, and 3-modules, respectively. Gene ontology analysis revealed significant enrichment of dynamic modules in pathways including translation, protein metabolic/catabolic processes, protein complex assembly, and endocytosis. Candidate CFTR effectors identified in the analysis included CHURC1, GZF1, and RPL15, and siRNA-mediated knockdown of these genes partially restored CFTR-dependent transepithelial chloride current to ΔF508-CFBE cells.

CONCLUSIONS

The ability of the M-module to identify dynamic modules involved in ΔF508 rescue provides a novel approach for studying CFTR biogenesis and identifying candidate suppressors of ΔF508.

A protein interaction landscape of breast cancer

[Figure: see text].

Inferring Functional Epigenetic Modules by Integrative Analysis of Multiple Heterogeneous Networks

Gene expression and methylation are critical biological processes for cells, and how to integrate these heterogeneous data has been extensively investigated, which is the foundation for revealing the underlying patterns of cancers. The vast majority of the current algorithms fuse gene methylation and expression into a network, failing to fully explore the relations and heterogeneity of them. To resolve these problems, in this study we define the epigenetic modules as a gene set whose members are co-methylated and co-expressed. To address the heterogeneity of data, we construct gene co-expression and co-methylation networks, respectively. In this case, the epigenetic module is characterized as a common module in multiple networks. Then, a non-negative matrix factorization-based algorithm that jointly clusters the co-expression and co-methylation networks is proposed for discovering the epigenetic modules (called Ep-jNMF). Ep-jNMF is more accurate than the baselines on the artificial data. Moreover, Ep-jNMF identifies more biologically meaningful modules. And the modules can predict the subtypes of cancers. These results indicate that Ep-jNMF is efficient for the integration of expression and methylation data.

Spatiotemporal 22q11.21 Protein Network Implicates DGCR8-Dependent MicroRNA Biogenesis as a Risk for Late-Fetal Cortical Development in Psychiatric Diseases

The chromosome 22q11.21 copy number variant (CNV) is a vital risk factor that can be a genetic predisposition to neurodevelopmental disorders (NDD). As the 22q11.21 CNV affects multiple genes, causal disease genes and mechanisms affected are still poorly understood. Thus, we aimed to identify the most impactful 22q11.21 CNV genes and the potential impacted human brain regions, developmental stages and signaling pathways. We constructed the spatiotemporal dynamic networks of 22q11.21 CNV genes using the brain developmental transcriptome and physical protein–protein interactions. The affected brain regions, developmental stages, driver genes and pathways were subsequently investigated via integrated bioinformatics analysis. As a result, we first identified that 22q11.21 CNV genes affect the cortical area mainly during late fetal periods. Interestingly, we observed that connections between a driver gene, DGCR8, and its interacting partners, MECP2 and CUL3, also network hubs, only existed in the network of the late fetal period within the cortical region, suggesting their functional specificity during brain development. We also confirmed the physical interaction result between DGCR8 and CUL3 by liquid chromatography-tandem mass spectrometry. In conclusion, our results could suggest that the disruption of DGCR8-dependent microRNA biogenesis plays a vital role in NDD for late fetal cortical development.

PRER: A patient representation with pairwise relative expression of proteins on biological networks

Changes in protein and gene expression levels are often used as features in predictive modeling such as survival prediction. A common strategy to aggregate information contained in individual proteins is to integrate the expression levels with the biological networks. In this work, we propose a novel patient representation where we integrate proteins’ expression levels with the protein-protein interaction (PPI) networks: Patient representation with PRER (Pairwise Relative Expressions with Random walks). PRER captures the dysregulation patterns of proteins based on the neighborhood of a protein in the PPI network. Specifically, PRER computes a feature vector for a patient by comparing the source protein’s expression level with other proteins’ levels that are within its neighborhood. The neighborhood of the source protein is derived by biased random-walk strategy on the network. We test PRER’s performance in survival prediction task in 10 different cancers using random forest survival models. PRER yields a statistically significant predictive performance in 9 out of 10 cancers when compared to the same model trained with features based on individual protein expressions. Furthermore, we identified the pairs of proteins that their interactions are predictive of patient survival but their individual expression levels are not. The set of identified relations provides a valuable collection of protein biomarkers with high prognostic value. PRER can be used for other complex diseases and prediction tasks that use molecular expression profiles as input. PRER is freely available at: https://github.com/hikuru/PRER.

Cancer remains to be one of the most prevalent and challenging diseases to treat. Cancer is a complex disease with several disrupted molecular mechanisms at play. The protein expression level is a fundamental indicator of how the molecular mechanisms are altered in each tumor. Predicting patient survival based on the changes is essential for understanding the cancer mechanisms and arriving at patient-specific treatment plans. For this task, existing machine learning models are used, such as random survival forest, which requires a feature-based representation of each patient based on her tumors. Most of these models use the individual molecular quantities of the tumors. However, cancer is a complex disease in which molecular mechanisms are dysregulated in various ways. In this work, we present a new patient representation scheme in which we integrate each tumor’s protein expression levels with their neighboring proteins’ expression levels in a protein-protein interaction network to capture patient-specific dysregulation patterns. Our results suggest that proteins’ relative expressions are more predictive than their individual expressions. We also analyze which of the protein interactions are more predictive of patient survival. The identified set of important protein interactions can be potentially used for cancer prognosis.

A Pipeline for Integrated Theory and Data-Driven Modeling of Biomedical Data

Genome sequencing technologies have the potential to transform clinical decision making and biomedical research by enabling high-throughput measurements of the genome at a granular level. However, to truly understand mechanisms of disease and predict the effects of medical interventions, high-throughput data must be integrated with demographic, phenotypic, environmental, and behavioral data from individuals. Further, effective knowledge discovery methods must infer relationships between these data types. We recently proposed a pipeline (CausalMGM) to achieve this. CausalMGM uses probabilistic graphical models to infer the relationships between variables in the data; however, CausalMGM's graphical structure learning algorithm can only handle small datasets efficiently. We propose a new methodology (piPref-Div) that selects the most informative variables for CausalMGM, enabling it to scale. We validate the efficacy of piPref-Div against other feature selection methods and demonstrate how the use of the full pipeline improves breast cancer outcome prediction and provides biologically interpretable views of gene expression data.

Stepwise target controllability identifies dysregulations of macrophage networks in multiple sclerosis

Identifying the nodes able to drive the state of a network is crucial to understand, and eventually control, biological systems. Despite recent advances, such identification remains difficult because of the huge number of equivalent controllable configurations, even in relatively simple networks. Based on the evidence that in many applications it is essential to test the ability of individual nodes to control a specific target subset, we develop a fast and principled method to identify controllable driver-target configurations in sparse and directed networks. We demonstrate our approach on simulated networks and experimental gene networks to characterize macrophage dysregulation in human subjects with multiple sclerosis.

Characterization of the Survival Influential Genes in Carcinogenesis

The genes influencing cancer patient mortality have been studied by survival analysis for many years. However, most studies utilized them only to support their findings associated with patient prognosis: their roles in carcinogenesis have not yet been revealed. Herein, we applied an in silico approach, integrating the Cox regression model with effect size estimated by the Monte Carlo algorithm, to screen survival-influential genes in more than 6000 tumor samples across 16 cancer types. We observed that the survival-influential genes had cancer-dependent properties. Moreover, the functional modules formed by the harmful genes were consistently associated with cell cycle in 12 out of the 16 cancer types and pan-cancer, showing that dysregulation of the cell cycle could harm patient prognosis in cancer. The functional modules formed by the protective genes are more diverse in cancers; the most prevalent functions are relevant for immune response, implying that patients with different cancer types might develop different mechanisms against carcinogenesis. We also identified a harmful set of 10 genes, with potential as prognostic biomarkers in pan-cancer. Briefly, our results demonstrated that the survival-influential genes could reveal underlying mechanisms in carcinogenesis and might provide clues for developing therapeutic targets for cancers.

Protein Interaction Domains: Structural Features and Drug Discovery Applications (Part 2)

BACKGROUND

Proteins present a modular organization made up of several domains. Apart from the domains playing catalytic functions, many others are crucial to recruit interactors. The latter domains can be defined as "PIDs" (Protein Interaction Domains) and are responsible for pivotal outcomes in signal transduction and a certain array of normal physiological and disease-related pathways. Targeting such PIDs with small molecules and peptides able to modulate their interaction networks, may represent a valuable route to discover novel therapeutics.

OBJECTIVE

This work represents a continuation of a very recent review describing PIDs able to recognize post-translationally modified peptide segments. On the contrary, the second part concerns with PIDs that interact with simple peptide sequences provided with standard amino acids.

METHODS

Crucial structural information on different domain subfamilies and their interactomes was gained by a wide search in different online available databases (including the PDB (Protein Data Bank), the Pfam (Protein family), and the SMART (Simple Modular Architecture Research Tool)). Pubmed was also searched to explore the most recent literature related to the topic.

RESULTS AND CONCLUSION

PIDs are multifaceted: they have all diverse structural features and can recognize several consensus sequences. PIDs can be linked to different diseases onset and progression, like cancer or viral infections and find applications in the personalized medicine field. Many efforts have been centered on peptide/peptidomimetic inhibitors of PIDs mediated interactions but much more work needs to be conducted to improve drug-likeness and interaction affinities of identified compounds.

Integrative genomics analysis identifies five promising genes implicated in insomnia risk based on multiple omics datasets

In recent decades, many genome-wide association studies on insomnia have reported numerous genes harboring multiple risk variants. Nevertheless, the molecular functions of these risk variants conveying risk to insomnia are still ill-studied. In the present study, we integrated GWAS summary statistics (N=386,533) with two independent brain expression quantitative trait loci (eQTL) datasets (N=329) to determine whether expression-associated SNPs convey risk to insomnia. Furthermore, we applied numerous bioinformatics analyses to highlight promising genes associated with insomnia risk. By using Sherlock integrative analysis, we detected 449 significant insomnia-associated genes in the discovery stage. These identified genes were significantly overrepresented in six biological pathways including Huntington’s disease (P=5.58 × 10⁻⁵), Alzheimer’s disease (P=5.58 × 10⁻⁵), Parkinson’s disease (P=6.34 × 10⁻⁵), spliceosome (P=1.17 × 10⁻⁴), oxidative phosphorylation (P=1.09 × 10⁻⁴), and wnt signaling pathways (P=2.07 × 10⁻⁴). Further, five of these identified genes were replicated in an independent brain eQTL dataset. Through a PPI network analysis, we found that there existed highly functional interactions among these five identified genes. Three genes of LDHA (P=0.044), DALRD3 (P=5.0 × 10⁻⁵), and HEBP2 (P=0.032) showed significantly lower expression level in brain tissues of insomnic patients than that in controls. In addition, the expression levels of these five genes showed prominently dynamic changes across different time points between behavioral states of sleep and sleep deprivation in mice brain cortex. Together, the evidence of the present study strongly suggested that these five identified genes may represent candidate genes and contributed risk to the etiology of insomnia.

Computational study of zebrafish immune-targeted microarray data for prediction of preventive drug candidates

Viral hemorrhagic septicemia virus (VHSV) is a rhabdovirus reported to cause economic loss in fish farms. Because of the lack of adequate preventative treatments, the identification of multipath genes involved in VHS infection might be an alternative to explore the possibility of using drugs for the seasonal prevention of this fish disease. We propose labeling a category of drug molecules by further classification and interpretation of the Drug Gene Interaction Database using gene ontology and Kyoto Encyclopedia of Genes and Genomes enrichment scores. The study investigated disease networks of up-and down-regulated genes to find those with high interaction as substantial genes in pathways among the different disease networks. We prioritized these genes based on their relationship to those associated with VHS infection in the context of human protein-protein interaction networks and disease pathways. Among the 29 genes as potential drug targets, nine were selected as promising druggable genes (ERBB2, FGFR3, ITGA2B, MAP2K1, NGF, NTRK1, PDGFRA, SCN2B, and SERPINC1). PDGFRA is the most important druggable up-and down-regulated gene and is considered an important gene in the IMATINIB pathway. This study findings indicate a promising approach for drug target prediction for VHS treatment, which might be useful for disease therapeutics.

Differential Network Analysis: A Statistical Perspective

Networks effectively capture interactions among components of complex systems, and have thus become a mainstay in many scientific disciplines. Growing evidence, especially from biology, suggest that networks undergo changes over time, and in response to external stimuli. In biology and medicine, these changes have been found to be predictive of complex diseases. They have also been used to gain insight into mechanisms of disease initiation and progression. Primarily motivated by biological applications, this article provides a review of recent statistical machine learning methods for inferring networks and identifying changes in their structures.

A mixture model to detect edges in sparse co-expression graphs with an application for comparing breast cancer subtypes

We develop a method to recover a gene network's structure from co-expression data, measured in terms of normalized Pearson's correlation coefficients between gene pairs. We treat these co-expression measurements as weights in the complete graph in which nodes correspond to genes. To decide which edges exist in the gene network, we fit a three-component mixture model such that the observed weights of 'null edges' follow a normal distribution with mean 0, and the non-null edges follow a mixture of two lognormal distributions, one for positively- and one for negatively-correlated pairs. We show that this so-called L2 N mixture model outperforms other methods in terms of power to detect edges, and it allows to control the false discovery rate. Importantly, our method makes no assumptions about the true network structure. We demonstrate our method, which is implemented in an R package called edgefinder, using a large dataset consisting of expression values of 12,750 genes obtained from 1,616 women. We infer the gene network structure by cancer subtype, and find insightful subtype characteristics. For example, we find thirteen pathways which are enriched in each of the cancer groups but not in the Normal group, with two of the pathways associated with autoimmune diseases and two other with graft rejection. We also find specific characteristics of different breast cancer subtypes. For example, the Luminal A network includes a single, highly connected cluster of genes, which is enriched in the human diseases category, and in the Her2 subtype network we find a distinct, and highly interconnected cluster which is uniquely enriched in drug metabolism pathways.

A network-based machine-learning framework to identify both functional modules and disease genes

Disease gene identification is a critical step towards uncovering the molecular mechanisms of diseases and systematically investigating complex disease phenotypes. Despite considerable efforts to develop powerful computing methods, candidate gene identification remains a severe challenge owing to the connectivity of an incomplete interactome network, which hampers the discovery of true novel candidate genes. We developed a network-based machine-learning framework to identify both functional modules and disease candidate genes. In this framework, we designed a semi-supervised non-negative matrix factorization model to obtain the functional modules related to the diseases and genes. Of note, we proposed a disease gene-prioritizing method called MapGene that integrates the correlations from both functional modules and network closeness. Our framework identified a set of functional modules with highly functional homogeneity and close gene interactions. Experiments on a large-scale benchmark dataset showed that MapGene performs significantly better than the state-of-the-art algorithms. Further analysis demonstrates MapGene can effectively relieve the impact of the incompleteness of interactome networks and obtain highly reliable rankings of candidate genes. In addition, disease cases on Parkinson's disease and diabetes mellitus confirmed the generalization of MapGene for novel candidate gene identification. This work proposed, for the first time, an integrated computing framework to predict both functional modules and disease candidate genes. The methodology and results support that our framework has the potential to help discover underlying functional modules and reliable candidate genes in human disease.

Network Analysis of Integrin Adhesion Complexes

Cell-surface adhesion receptors mediate interactions with the extracellular matrix (ECM) to control many fundamental aspects of cell behavior, including cell migration, survival, and proliferation. Integrin adhesion receptors recruit structural and signaling proteins to form multimolecular adhesion complexes that link the plasma membrane to the actomyosin cytoskeleton. The assembly and turnover of adhesion complexes are tightly regulated, governed in part by the networks of physical protein interactions and functional signaling associations between components of the adhesome. Proteomic profiling of adhesion complexes has begun to reveal their molecular complexity and diversity. To interrogate the composition of cell-ECM adhesions, we detail herein an approach for the network analysis of adhesion complex proteomes. Integration of these proteomic data with adhesome databases in the context of predicted protein interactions enables the mapping of experimentally defined adhesion complex networks. Computational analysis of resultant network models can identify subnetworks of putative functionally linked adhesion protein communities. This approach provides a framework to predict functional adhesion protein relationships and generate new mechanistic hypotheses for further experimental testing.

Differential Co-Expression Analyses Allow the Identification of Critical Signalling Pathways Altered during Tumour Transformation and Progression

Biological systems respond to perturbations through the rewiring of molecular interactions, organised in gene regulatory networks (GRNs). Among these, the increasingly high availability of transcriptomic data makes gene co-expression networks the most exploited ones. Differential co-expression networks are useful tools to identify changes in response to an external perturbation, such as mutations predisposing to cancer development, and leading to changes in the activity of gene expression regulators or signalling. They can help explain the robustness of cancer cells to perturbations and identify promising candidates for targeted therapy, moreover providing higher specificity with respect to standard co-expression methods. Here, we comprehensively review the literature about the methods developed to assess differential co-expression and their applications to cancer biology. Via the comparison of normal and diseased conditions and of different tumour stages, studies based on these methods led to the definition of pathways involved in gene network reorganisation upon oncogenes’ mutations and tumour progression, often converging on immune system signalling. A relevant implementation still lagging behind is the integration of different data types, which would greatly improve network interpretability. Most importantly, performance and predictivity evaluation of the large variety of mathematical models proposed would urgently require experimental validations and systematic comparisons. We believe that future work on differential gene co-expression networks, complemented with additional omics data and experimentally tested, will considerably improve our insights into the biology of tumours.

Effects of active farnesoid X receptor on GLUTag enteroendocrine L cells

Activated transcription factor (TF) farnesoid X receptor (FXR) represses glucagon-like peptide-1 (GLP-1) secretion in enteroendocrine L cells. This, in turn, reduces insulin secretion, which is triggered when β cells bind GLP-1. Preventing FXR activation could boost GLP-1 production and insulin secretion. Yet, FXR's broader role in L cell biology still lacks understanding. Here, we show that FXR is a multifaceted TF in L cells using proteomics and gene expression data generated on GLUTag L cells. Most striking, 252 proteins regulated upon glucose stimulation have their abundances neutralized upon FXR activation. Mitochondrial repression or glucose import block are likely mechanisms of this. Further, FXR physically targets bile acid metabolism proteins, growth factors and other TFs, regulates ChREBP, while extensive text-mining found 30 FXR-regulated proteins to be well-known in L cell biology. Taken together, this outlines FXR as a powerful TF, where GLP-1 secretion block is just one of many downstream effects.

Finding disease modules for cancer and COVID-19 in gene co-expression networks with the Core&Peel method

Genes are organized in functional modules (or pathways), thus their action and their dysregulation in diseases may be better understood by the identification of the modules most affected by the disease (aka disease modules, or active subnetworks). We describe how an algorithm based on the Core&Peel method is used to detect disease modules in co-expression networks of genes. We first validate Core&Peel for the general task of functional module detection by comparison with 42 methods participating in the Disease Module Identification DREAM challenge. Next, we use four specific disease test cases (colorectal cancer, prostate cancer, asthma, and rheumatoid arthritis), four state-of-the-art algorithms (ModuleDiscoverer, Degas, KeyPathwayMiner, and ClustEx), and several pathway databases to validate the proposed algorithm. Core&Peel is the only method able to find significant associations of the predicted disease module with known validated relevant pathways for all four diseases. Moreover, for the two cancer datasets, Core&Peel detects further eight relevant pathways not discovered by the other methods used in the comparative analysis. Finally, we apply Core&Peel and other methods to explore the transcriptional response of human cells to SARS-CoV-2 infection, finding supporting evidence for drug repositioning efforts at a pre-clinical level.

Integrated genomics analysis highlights important SNPs and genes implicated in moderate-to-severe asthma based on GWAS and eQTL datasets

Background

Severe asthma is a chronic disease contributing to disproportionate disease morbidity and mortality. From the year of 2007, many genome-wide association studies (GWAS) have documented a large number of asthma-associated genetic variants and related genes. Nevertheless, the molecular mechanism of these identified variants involved in asthma or severe asthma risk remains largely unknown.

Methods

In the current study, we systematically integrated 3 independent expression quantitative trait loci (eQTL) data (N = 1977) and a large-scale GWAS summary data of moderate-to-severe asthma (N = 30,810) by using the Sherlock Bayesian analysis to identify whether expression-related variants contribute risk to severe asthma. Furthermore, we performed various bioinformatics analyses, including pathway enrichment analysis, PPI network enrichment analysis, in silico permutation analysis, DEG analysis and co-expression analysis, to prioritize important genes associated with severe asthma.

Results

In the discovery stage, we identified 1129 significant genes associated with moderate-to-severe asthma by using the Sherlock Bayesian analysis. Two hundred twenty-eight genes were prominently replicated by using MAGMA gene-based analysis. These 228 replicated genes were enriched in 17 biological pathways including antigen processing and presentation (Corrected P = 4.30 × 10^− 6), type I diabetes mellitus (Corrected P = 7.09 × 10^− 5), and asthma (Corrected P = 1.72 × 10^− 3). With the use of a series of bioinformatics analyses, we highlighted 11 important genes such as GNGT2, TLR6, and TTC19 as authentic risk genes associated with moderate-to-severe/severe asthma. With respect to GNGT2, there were 3 eSNPs of rs17637472 (P_eQTL = 2.98 × 10^− 8 and P_GWAS = 3.40 × 10^− 8), rs11265180 (P_eQTL = 6.0 × 10^− 6 and P_GWAS = 1.99 × 10^− 3), and rs1867087 (P_eQTL = 1.0 × 10^− 4 and P_GWAS = 1.84 × 10^− 5) identified. In addition, GNGT2 is significantly expressed in severe asthma compared with mild-moderate asthma (P = 0.045), and Gngt2 shows significantly distinct expression patterns between vehicle and various glucocorticoids (Anova P = 1.55 × 10^− 6).

Conclusions

Our current study provides multiple lines of evidence to support that these 11 identified genes as important candidates implicated in the pathogenesis of severe asthma.

Designing a Network Proximity-Based Drug Repurposing Strategy for COVID-19

The ongoing COVID-19 pandemic still requires fast and effective efforts from all fronts, including epidemiology, clinical practice, molecular medicine, and pharmacology. A comprehensive molecular framework of the disease is needed to better understand its pathological mechanisms, and to design successful treatments able to slow down and stop the impressive pace of the outbreak and harsh clinical symptomatology, possibly via the use of readily available, off-the-shelf drugs. This work engages in providing a wider picture of the human molecular landscape of the SARS-CoV-2 infection via a network medicine approach as the ground for a drug repurposing strategy. Grounding on prior knowledge such as experimentally validated host proteins known to be viral interactors, tissue-specific gene expression data, and using network analysis techniques such as network propagation and connectivity significance, the host molecular reaction network to the viral invasion is explored and exploited to infer and prioritize candidate target genes, and finally to propose drugs to be repurposed for the treatment of COVID-19. Ranks of potential target genes have been obtained for coherent groups of tissues/organs, potential and distinct sites of interaction between the virus and the organism. The normalization and the aggregation of the different scores allowed to define a preliminary, restricted list of genes candidates as pharmacological targets for drug repurposing, with the aim of contrasting different phases of the virus infection and viral replication cycle.

PALMER: improving pathway annotation based on the biomedical literature mining with a constrained latent block model

Background

In systems biology, it is of great interest to identify previously unreported associations between genes. Recently, biomedical literature has been considered as a valuable resource for this purpose. While classical clustering algorithms have popularly been used to investigate associations among genes, they are not tuned for the literature mining data and are also based on strong assumptions, which are often violated in this type of data. For example, these approaches often assume homogeneity and independence among observations. However, these assumptions are often violated due to both redundancies in functional descriptions and biological functions shared among genes. Latent block models can be alternatives in this case but they also often show suboptimal performances, especially when signals are weak. In addition, they do not allow to utilize valuable prior biological knowledge, such as those available in existing databases.

Results

In order to address these limitations, here we propose PALMER, a constrained latent block model that allows to identify indirect relationships among genes based on the biomedical literature mining data. By automatically associating relevant Gene Ontology terms, PALMER facilitates biological interpretation of novel findings without laborious downstream analyses. PALMER also allows researchers to utilize prior biological knowledge about known gene-pathway relationships to guide identification of gene–gene associations. We evaluated PALMER with simulation studies and applications to studies of pathway-modulating genes relevant to cancer signaling pathways, while utilizing biological pathway annotations available in the KEGG database as prior knowledge.

Conclusions

We showed that PALMER outperforms traditional latent block models and it provides reliable identification of novel gene–gene associations by utilizing prior biological knowledge, especially when signals are weak in the biomedical literature mining dataset. We believe that PALMER and its relevant user-friendly software will be powerful tools that can be used to improve existing pathway annotations and identify novel pathway-modulating genes.

Robust edge-based biomarker discovery improves prediction of breast cancer metastasis

Background

The abundance of molecular profiling of breast cancer tissues entailed active research on molecular marker-based early diagnosis of metastasis. Recently there is a surging interest in combining gene expression with gene networks such as protein-protein interaction (PPI) network, gene co-expression (CE) network and pathway information to identify robust and accurate biomarkers for metastasis prediction, reflecting the common belief that cancer is a systems biology disease. However, controversy exists in the literature regarding whether network markers are indeed better features than genes alone for predicting as well as understanding metastasis. We believe much of the existing results may have been biased by the overly complicated prediction algorithms, unfair evaluation, and lack of rigorous statistics. In this study, we propose a simple approach to use network edges as features, based on two types of networks respectively, and compared their prediction power using three classification algorithms and rigorous statistical procedure on one of the largest datasets available. To detect biomarkers that are significant for the prediction and to compare the robustness of different feature types, we propose an unbiased and novel procedure to measure feature importance that eliminates the potential bias from factors such as different sample size, number of features, as well as class distribution.

Results

Experimental results reveal that edge-based feature types consistently outperformed gene-based feature type in random forest and logistic regression models under all performance evaluation metrics, while the prediction accuracy of edge-based support vector machine (SVM) model was poorer, due to the larger number of edge features compared to gene features and the lack of feature selection in SVM model. Experimental results also show that edge features are much more robust than gene features and the top biomarkers from edge feature types are statistically more significantly enriched in the biological processes that are well known to be related to breast cancer metastasis.

Conclusions

Overall, this study validates the utility of edge features as biomarkers but also highlights the importance of carefully designed experimental procedures in order to achieve statistically reliable comparison results.

A large accessory protein interactome is rewired across environments

To characterize how protein-protein interaction (PPI) networks change, we quantified the relative PPI abundance of 1.6 million protein pairs in the yeast Saccharomyces cerevisiae across nine growth conditions, with replication, for a total of 44 million measurements. Our multi-condition screen identified 13,764 pairwise PPIs, a threefold increase over PPIs identified in one condition. A few 'immutable' PPIs are present across all conditions, while most 'mutable' PPIs are rarely observed. Immutable PPIs aggregate into highly connected 'core' network modules, with most network remodeling occurring within a loosely connected 'accessory' module. Mutable PPIs are less likely to co-express, co-localize, and be explained by simple mass action kinetics, and more likely to contain proteins with intrinsically disordered regions, implying that environment-dependent association and binding is critical to cellular adaptation. Our results show that protein interactomes are larger than previously thought and contain highly dynamic regions that reorganize to drive or respond to cellular changes.