Systematic Functional Annotation and Visualization of Biological Networks

SDN2GO: An Integrated Deep Learning Model for Protein Function Prediction

The assignment of function to proteins at a large scale is essential for understanding the molecular mechanism of life. However, only a very small percentage of the more than 179 million proteins in UniProtKB have Gene Ontology (GO) annotations supported by experimental evidence. In this paper, we proposed an integrated deep-learning-based classification model, named SDN2GO, to predict protein functions. SDN2GO applies convolutional neural networks to learn and extract features from sequences, protein domains, and known PPI networks, and then utilizes a weight classifier to integrate these features and achieve accurate predictions of GO terms. We constructed the training set and the independent test set according to the time-delayed principle of the Critical Assessment of Function Annotation (CAFA) and compared it with two highly competitive methods and the classic BLAST method on the independent test set. The results show that our method outperforms others on each sub-ontology of GO. We also investigated the performance of using protein domain information. We learned from the Natural Language Processing (NLP) to process domain information and pre-trained a deep learning sub-model to extract the comprehensive features of domains. The experimental results demonstrate that the domain features we obtained are much improved the performance of our model. Our deep learning models together with the data pre-processing scripts are publicly available as an open source software at https://github.com/Charrick/SDN2GO.

A reference map of the human binary protein interactome

Global insights into cellular organization and genome function require comprehensive understanding of the interactome networks that mediate genotype-phenotype relationships^1,2. Here, we present a human “all-by-all” reference interactome map of human binary protein interactions, or “HuRI”. With ~53,000 high-quality protein-protein interactions (PPIs), HuRI has approximately four times more such interactions than high-quality curated interactions from small-scale studies. Integrating HuRI with genome³, transcriptome⁴, and proteome⁵ data enables the study of cellular function within most physiological or pathological cellular contexts. We demonstrate the utility of HuRI in identifying specific subcellular roles of PPIs. Inferred tissue-specific networks reveal general principles for the formation of cellular context-specific functions and elucidate potential molecular mechanisms underlying tissue-specific phenotypes of Mendelian diseases. HuRI represents a systematic proteome-wide reference linking genomic variation to phenotypic outcomes.

Prospects and challenges of multi-omics data integration in toxicology

Exposure of cells or organisms to chemicals can trigger a series of effects at the regulatory pathway level, which involve changes of levels, interactions, and feedback loops of biomolecules of different types. A single-omics technique, e.g., transcriptomics, will detect biomolecules of one type and thus can only capture changes in a small subset of the biological cascade. Therefore, although applying single-omics analyses can lead to the identification of biomarkers for certain exposures, they cannot provide a systemic understanding of toxicity pathways or adverse outcome pathways. Integration of multiple omics data sets promises a substantial improvement in detecting this pathway response to a toxicant, by an increase of information as such and especially by a systemic understanding. Here, we report the findings of a thorough evaluation of the prospects and challenges of multi-omics data integration in toxicological research. We review the availability of such data, discuss options for experimental design, evaluate methods for integration and analysis of multi-omics data, discuss best practices, and identify knowledge gaps. Re-analyzing published data, we demonstrate that multi-omics data integration can considerably improve the confidence in detecting a pathway response. Finally, we argue that more data need to be generated from studies with a multi-omics-focused design, to define which omics layers contribute most to the identification of a pathway response to a toxicant.

Systematic genetics and single-cell imaging reveal widespread morphological pleiotropy and cell-to-cell variability

Our ability to understand the genotype-to-phenotype relationship is hindered by the lack of detailed understanding of phenotypes at a single-cell level. To systematically assess cell-to-cell phenotypic variability, we combined automated yeast genetics, high-content screening and neural network-based image analysis of single cells, focussing on genes that influence the architecture of four subcellular compartments of the endocytic pathway as a model system. Our unbiased assessment of the morphology of these compartments-endocytic patch, actin patch, late endosome and vacuole-identified 17 distinct mutant phenotypes associated with ~1,600 genes (~30% of all yeast genes). Approximately half of these mutants exhibited multiple phenotypes, highlighting the extent of morphological pleiotropy. Quantitative analysis also revealed that incomplete penetrance was prevalent, with the majority of mutants exhibiting substantial variability in phenotype at the single-cell level. Our single-cell analysis enabled exploration of factors that contribute to incomplete penetrance and cellular heterogeneity, including replicative age, organelle inheritance and response to stress.

A genome-wide screen identifies genes that suppress the accumulation of spontaneous mutations in young and aged yeast cells

To ensure proper transmission of genetic information, cells need to preserve and faithfully replicate their genome, and failure to do so leads to genome instability, a hallmark of both cancer and aging. Defects in genes involved in guarding genome stability cause several human progeroid syndromes, and an age-dependent accumulation of mutations has been observed in different organisms, from yeast to mammals. However, it is unclear whether the spontaneous mutation rate changes during aging and whether specific pathways are important for genome maintenance in old cells. We developed a high-throughput replica-pinning approach to screen for genes important to suppress the accumulation of spontaneous mutations during yeast replicative aging. We found 13 known mutation suppression genes, and 31 genes that had no previous link to spontaneous mutagenesis, and all acted independently of age. Importantly, we identified PEX19, encoding an evolutionarily conserved peroxisome biogenesis factor, as an age-specific mutation suppression gene. While wild-type and pex19Δ young cells have similar spontaneous mutation rates, aged cells lacking PEX19 display an elevated mutation rate. This finding suggests that functional peroxisomes may be important to preserve genome integrity specifically in old cells.

tmap: an integrative framework based on topological data analysis for population-scale microbiome stratification and association studies

Untangling the complex variations of microbiome associated with large-scale host phenotypes or environment types challenges the currently available analytic methods. Here, we present tmap, an integrative framework based on topological data analysis for population-scale microbiome stratification and association studies. The performance of tmap in detecting nonlinear patterns is validated by different scenarios of simulation, which clearly demonstrate its superiority over the most commonly used methods. Application of tmap to several population-scale microbiomes extensively demonstrates its strength in revealing microbiome-associated host or environmental features and in understanding the systematic interrelations among their association patterns. tmap is available at https://github.com/GPZ-Bioinfo/tmap.

Genetic interaction networks mediate individual statin drug response in Saccharomyces cerevisiae

Eukaryotic genetic interaction networks (GINs) are extensively described in the Saccharomyces cerevisiae S288C model using deletion libraries, yet being limited to this one genetic background, not informative to individual drug response. Here we created deletion libraries in three additional genetic backgrounds. Statin response was probed with five queries against four genetic backgrounds. The 20 resultant GINs representing drug-gene and gene-gene interactions were not conserved by functional enrichment, hierarchical clustering, and topology-based community partitioning. An unfolded protein response (UPR) community exhibited genetic background variation including different betweenness genes that were network bottlenecks, and we experimentally validated this UPR community via measurements of the UPR that were differentially activated and regulated in statin-resistant strains relative to the statin-sensitive S288C background. These network analyses by topology and function provide insight into the complexity of drug response influenced by genetic background.

Functional impact of microRNA regulation in models of extreme stress adaptation

When confronted with severe environmental stress, some animals are able to undergo a substantial reorganization of their cellular environment that enables long-term survival. One molecular mechanism of adaptation that has received considerable attention in recent years has been the action of reversible transcriptome regulation by microRNA. The implementation of new computational and high-throughput experimental approaches has started to uncover the vital contributions of microRNA towards stress adaptation. Indeed, recent studies have suggested that microRNA may have a major regulatory influence over a number of cellular processes that are essential to prolonged environmental stress survival. To date, a number of studies have highlighted the role of microRNA in the regulation of a metabolically depressed state, documenting stress-responsive microRNA expression during mammalian hibernation, frog and insect freeze tolerance, and turtle and marine snail anoxia tolerance. These studies collectively indicate a conserved principle of microRNA stress response across phylogeny. As we are on the verge of dissecting the role of microRNA in environmental stress adaptation, this review summarizes recent research advances and the hallmark expression patterns that facilitate stress survival.

Mitochondrial genomic variation drives differential nuclear gene expression in discrete regions of Drosophila gene and protein interaction networks

BACKGROUND

Mitochondria perform many key roles in their eukaryotic hosts, from integrating signaling pathways through to modulating whole organism phenotypes. The > 1 billion years of nuclear and mitochondrial gene co-evolution has necessitated coordinated expression of gene products from both genomes that maintain mitochondrial, and more generally, eukaryotic cellular function. How mitochondrial DNA (mtDNA) variation modifies host fitness has proved a challenging question but has profound implications for evolutionary and medical genetics. In Drosophila, we have previously shown that recently diverged mtDNA haplotypes within-species can have more impact on organismal phenotypes than older, deeply diverged haplotypes from different species. Here, we tested the effects of mtDNA haplotype variation on gene expression in Drosophila under standardized conditions. Using the Drosophila Genetic Reference Panel (DGRP), we constructed a panel of mitonuclear genotypes that consists of factorial variation in nuclear and mtDNA genomes, with mtDNAs originating in D. melanogaster (2x haplotypes) and D. simulans (2x haplotypes).

RESULTS

We show that mtDNA haplotype variation unequivocally alters nuclear gene expression in both females and males, and mitonuclear interactions are pervasive modifying factors for gene expression. There was appreciable overlap between the sexes for mtDNA-sensitive genes, and considerable transcriptional variation attributed to particular mtDNA contrasts. These genes are generally found in low-connectivity gene co-expression networks, occur in gene clusters along chromosomes, are often flanked by non-coding RNA, and are under-represented among housekeeping genes. Finally, we identify the giant (gt) transcription factor motif as a putative regulatory sequence associated with mtDNA-sensitive genes.

CONCLUSIONS

There are predictive conditions for nuclear genes that are influenced by mtDNA variation.

Replication timing networks reveal a link between transcription regulatory circuits and replication timing control

DNA replication occurs in a defined temporal order known as the replication timing (RT) program and is regulated during development, coordinated with 3D genome organization and transcriptional activity. However, transcription and RT are not sufficiently coordinated to predict each other, suggesting an indirect relationship. Here, we exploit genome-wide RT profiles from 15 human cell types and intermediate differentiation stages derived from human embryonic stem cells to construct different types of RT regulatory networks. First, we constructed networks based on the coordinated RT changes during cell fate commitment to create highly complex RT networks composed of thousands of interactions that form specific functional subnetwork communities. We also constructed directional regulatory networks based on the order of RT changes within cell lineages, and identified master regulators of differentiation pathways. Finally, we explored relationships between RT networks and transcriptional regulatory networks (TRNs) by combining them into more complex circuitries of composite and bipartite networks. Results identified novel trans interactions linking transcription factors that are core to the regulatory circuitry of each cell type to RT changes occurring in those cell types. These core transcription factors were found to bind cooperatively to sites in the affected replication domains, providing provocative evidence that they constitute biologically significant directional interactions. Our findings suggest a regulatory link between the establishment of cell-type-specific TRNs and RT control during lineage specification.

A network of networks approach for modeling interconnected brain tissue-specific networks

Motivation

Recent sequence-based analyses have identified a lot of gene variants that may contribute to neurogenetic disorders such as autism spectrum disorder and schizophrenia. Several state-of-the-art network-based analyses have been proposed for mechanical understanding of genetic variants in neurogenetic disorders. However, these methods were mainly designed for modeling and analyzing single networks that do not interact with or depend on other networks, and thus cannot capture the properties between interdependent systems in brain-specific tissues, circuits and regions which are connected each other and affect behavior and cognitive processes.

Results

We introduce a novel and efficient framework, called a ‘Network of Networks’ approach, to infer the interconnectivity structure between multiple networks where the response and the predictor variables are topological information matrices of given networks. We also propose Graph-Oriented SParsE Learning, a new sparse structural learning algorithm for network data to identify a subset of the topological information matrices of the predictors related to the response. We demonstrate on simulated data that propose Graph-Oriented SParsE Learning outperforms existing kernel-based algorithms in terms of F-measure. On real data from human brain region-specific functional networks associated with the autism risk genes, we show that the ‘Network of Networks’ model provides insights on the autism-associated interconnectivity structure between functional interaction networks and a comprehensive understanding of the genetic basis of autism across diverse regions of the brain.

Availability and implementation

Our software is available from https://github.com/infinite-point/GOSPEL.

Supplementary information

Supplementary data are available at Bioinformatics online.

Measuring similarity between gene interaction profiles

Background

Gene and protein interaction data are often represented as interaction networks, where nodes stand for genes or gene products and each edge stands for a relationship between a pair of gene nodes. Commonly, that relationship within a pair is specified by high similarity between profiles (vectors) of experimentally defined interactions of each of the two genes with all other genes in the genome; only gene pairs that interact with similar sets of genes are linked by an edge in the network. The tight groups of genes/gene products that work together in a cell can be discovered by the analysis of those complex networks.

Results

We show that the choice of the similarity measure between pairs of gene vectors impacts the properties of networks and of gene modules detected within them. We re-analyzed well-studied data on yeast genetic interactions, constructed four genetic networks using four different similarity measures, and detected gene modules in each network using the same algorithm. The four networks induced different numbers of putative functional gene modules, and each similarity measure induced some unique modules. In an example of a putative functional connection suggested by comparing genetic interaction vectors, we predict a link between SUN-domain proteins and protein glycosylation in the endoplasmic reticulum.

Conclusions

The discovery of molecular modules in genetic networks is sensitive to the way of measuring similarity between profiles of gene interactions in a cell. In the absence of a formal way to choose the “best” measure, it is advisable to explore the measures with different mathematical properties, which may identify different sets of connections between genes.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-3024-x) contains supplementary material, which is available to authorized users.

Explorative analysis of the gene expression profile during liver regeneration of mouse: a microarray-based study

The liver is an amazing organ due to its powerful regenerative capacity. Although many studies on liver regeneration have been documented, the detailed mechanisms remain unclear. Two-third partial hepatectomy (PH) in rodents plays a crucial role in the study of liver regeneration. In this study, the time series data of gene expression during liver regeneration in mouse were analyzed using the gene set numbered GSE6998 in GEO. A variety of bioinformatics methods, including masigPro, Weighted Gene Co-expression Network Analysis (WGCNA), spatial analysis of functional enrichment (SAFE) and ingenuity canonical pathway analysis (IPA) were used to identify and compare the significantly changed pathways, potential upstream regulators and key genes during liver regeneration. Our study showed that liver regeneration in the mouse is a coordinated process, which cell-cycle-related progress are at the centre of the interaction network involved in liver regeneration. Several candidate upstream regulators including PPARA, NFE2L2, MAD1 and CNR1 and some key genes such as Cdk1, Plk1, Cdc20, Aurka, Racgap1, Cenpa, Rrm1, Rrm2 were identified. In conclusion, these findings could contribute to revealing the molecular mechanism of liver regeneration after PH, which could provide new ideas and treatment methods for regenerative medicine, oncological drug development and oncological treatment.

Network approaches to systems biology analysis of complex disease: integrative methods for multi-omics data

In the past decade, significant progress has been made in complex disease research across multiple omics layers from genome, transcriptome and proteome to metabolome. There is an increasing awareness of the importance of biological interconnections, and much success has been achieved using systems biology approaches. However, because of the typical focus on one single omics layer at a time, existing systems biology findings explain only a modest portion of complex disease. Recent advances in multi-omics data collection and sharing present us new opportunities for studying complex diseases in a more comprehensive fashion, and yet simultaneously create new challenges considering the unprecedented data dimensionality and diversity. Here, our goal is to review extant and emerging network approaches that can be applied across multiple biological layers to facilitate a more comprehensive and integrative multilayered omics analysis of complex diseases.

Exploratory Gene Ontology Analysis with Interactive Visualization

The Gene Ontology (GO) is a central resource for functional-genomics research. Scientists rely on the functional annotations in the GO for hypothesis generation and couple it with high-throughput biological data to enhance interpretation of results. At the same time, the sheer number of concepts (>30,000) and relationships (>70,000) presents a challenge: it can be difficult to draw a comprehensive picture of how certain concepts of interest might relate with the rest of the ontology structure. Here we present new visualization strategies to facilitate the exploration and use of the information in the GO. We rely on novel graphical display and software architecture that allow significant interaction. To illustrate the potential of our strategies, we provide examples from high-throughput genomic analyses, including chromatin immunoprecipitation experiments and genome-wide association studies. The scientist can also use our visualizations to identify gene sets that likely experience coordinated changes in their expression and use them to simulate biologically-grounded single cell RNA sequencing data, or conduct power studies for differential gene expression studies using our built-in pipeline. Our software and documentation are available at http://aegis.stanford.edu .

Global Genetic Networks and the Genotype-to-Phenotype Relationship

Genetic interactions identify combinations of genetic variants that impinge on phenotype. With whole-genome sequence information available for thousands of individuals within a species, a major outstanding issue concerns the interpretation of allelic combinations of genes underlying inherited traits. In this Review, we discuss how large-scale analyses in model systems have illuminated the general principles and phenotypic impact of genetic interactions. We focus on studies in budding yeast, including the mapping of a global genetic network. We emphasize how information gained from work in yeast translates to other systems, and how a global genetic network not only annotates gene function but also provides new insights into the genotype-to-phenotype relationship.

Towards a data-integrated cell

We are increasingly accumulating molecular data about a cell. The challenge is how to integrate them within a unified conceptual and computational framework enabling new discoveries. Hence, we propose a novel, data-driven concept of an integrated cell, iCell. Also, we introduce a computational prototype of an iCell, which integrates three omics, tissue-specific molecular interaction network types. We construct iCells of four cancers and the corresponding tissue controls and identify the most rewired genes in cancer. Many of them are of unknown function and cannot be identified as different in cancer in any specific molecular network. We biologically validate that they have a role in cancer by knockdown experiments followed by cell viability assays. We find additional support through Kaplan-Meier survival curves of thousands of patients. Finally, we extend this analysis to uncover pan-cancer genes. Our methodology is universal and enables integrative comparisons of diverse omics data over cells and tissues.

In Silico Oncology Drug Repositioning and Polypharmacology

Network-aided in silico approaches have been widely used for prediction of drug-target interactions and evaluation of drug safety to increase the clinical efficiency and productivity during drug discovery and development. Here we review the advances and new progress in this field and summarize the translational applications of several new network-aided in silico approaches we developed recently. In addition, we describe the detailed protocols for a network-aided drug repositioning infrastructure for identification of new targets for old drugs, failed drugs in clinical trials, and new chemical entities. These state-of-the-art network-aided in silico approaches have been used for the discovery and development of broad-acting and targeted clinical therapies for various complex diseases, in particular for oncology drug repositioning. In this chapter, the described network-aided in silico protocols are appropriate for target-centric drug repositioning to various complex diseases, but expertise is still necessary to perform the specific oncology projects based on the cancer targets of interest.

Comparison of gene expression in liver regeneration and hepatocellular carcinoma formation

Background

Liver -cell proliferation occurs in hepatocellular carcinoma (HCC) and liver regeneration (LR). The development and progression of HCC and LR have many similar molecular pathways with very different results. In simple terms, LR is a controllable process of organ recovery and function reconstruction, whereas liver cancer is uncontrollable. Do they share common key pathways and genes?

Methods

In this study, the dynamic transcriptome profile at ten time points (0, 2, 6, 12, 24, 30, 36, 72, 120, and 168 hours) during LR in rats after two-thirds hepatectomy and eight stages (normal, cirrhosis without HCC, cirrhosis, low-grade dysplastic, high-grade dysplastic, and very early, early advanced, and very advanced HCC) representing a stepwise carcinogenic process from preneoplastic lesions to end-stage HCC were analyzed in detail. A variety of bioinformatic methods, including MaSigPro, weighted gene-coexpression network analysis, and spatial analysis of functional enrichment, were used to analyze, elucidate, and compare similarities and differences between LR and HCC formation.

Results

Key biological processes and genes were identified. From the comparison, we found that cell proliferation and angiogenesis were the most significantly dysregulated processes shared by LR and HCC. The pattern of cell-proliferation-related gene expression in progression stage during LR is similar to the transition process from dysplasia to early-stage HCC. LR and HCC showed different expression patterns as a whole. Some key genes, including FYN, XPO1, FOXM1, EZH2, and NRF1, were identified as playing critical roles in both LR and HCC.

Conclusion

These findings could contribute to revealing the molecular mechanism of development and regulation mechanism of normal and abnormal proliferation, which could provide new ideas and treatment methods for regenerative medicine, oncological drug development, and oncological treatment.

DeepGO: predicting protein functions from sequence and interactions using a deep ontology-aware classifier

Motivation

A large number of protein sequences are becoming available through the application of novel high-throughput sequencing technologies. Experimental functional characterization of these proteins is time-consuming and expensive, and is often only done rigorously for few selected model organisms. Computational function prediction approaches have been suggested to fill this gap. The functions of proteins are classified using the Gene Ontology (GO), which contains over 40 000 classes. Additionally, proteins have multiple functions, making function prediction a large-scale, multi-class, multi-label problem.

Results

We have developed a novel method to predict protein function from sequence. We use deep learning to learn features from protein sequences as well as a cross-species protein–protein interaction network. Our approach specifically outputs information in the structure of the GO and utilizes the dependencies between GO classes as background information to construct a deep learning model. We evaluate our method using the standards established by the Computational Assessment of Function Annotation (CAFA) and demonstrate a significant improvement over baseline methods such as BLAST, in particular for predicting cellular locations.

Availability and implementation

Web server: http://deepgo.bio2vec.net, Source code: https://github.com/bio-ontology-research-group/deepgo

Supplementary information

Supplementary data are available at Bioinformatics online.

Simulated microgravity triggers characteristic morphology and stress response in Saccharomyces cerevisiae

Reduction of gravity results in changes in gene expression and morphology in the budding yeast Saccharomyces cerevisiae. We studied the genes responsible for the morphological changes induced by simulated microgravity (SMG) using the yeast morphology data. We comprehensively captured the features of the morphological changes in yeast cells cultured in SMG with CalMorph, a high-throughput image-processing system. Statistical analysis revealed that 95 of 501 morphological traits were significantly affected, which included changes in bud direction, the ratio of daughter to mother cell size, the random daughter cell shape, the large mother cell size, bright nuclei in the M phase, and the decrease in angle between two nuclei. We identified downregulated genes that impacted the morphological changes in conditions of SMG by focusing on each of the morphological features individually. Gene Ontology (GO)-enrichment analysis indicated that morphological changes under conditions of SMG were caused by cooperative downregulation of 103 genes annotated to six GO terms, which included cytoplasmic ribonucleoprotein granule, RNA elongation, mitotic cell cycle phase transition, nucleocytoplasmic transport, protein-DNA complex subunit organization, and RNA localization. P-body formation was also promoted under conditions of SMG. These results suggest that cooperative downregulation of multiple genes occurs in conditions of SMG.

Network-based genetic investigation of virulence-associated phenotypes in methicillin-resistant Staphylococcus aureus

Staphylococcus aureus is a gram-positive bacterium that causes a wide range of infections. Recently, the spread of methicillin-resistant S. aureus (MRSA) strains has seriously reduced antibiotic treatment options. Anti-virulence strategies, the objective of which is to target the virulence instead of the viability of the pathogen, have become widely accepted as a means of avoiding the emergence of new antibiotic-resistant strains. To increase the number of anti-virulence therapeutic options, it is necessary to identify as many novel virulence-associated genes as possible in MRSA. Co-functional networks have proved useful for mapping gene-to-phenotype associations in various organisms. Herein, we present StaphNet (www.inetbio.org/staphnet), a genome-scale co-functional network for an MRSA strain, S. aureus subsp. USA300_FPR3757. StaphNet, which was constructed by the integration of seven distinct types of genomics data within a Bayesian statistics framework, covers approximately 94% of the coding genome with a high degree of accuracy. We implemented a companion web server for network-based gene prioritization of the phenotypes of 31 different S. aureus strains. We demonstrated that StaphNet can effectively identify genes for virulence-associated phenotypes in MRSA. These results suggest that StaphNet can facilitate target discovery for the development of anti-virulence drugs to treat MRSA infection.

Prioritizing network communities

Uncovering modular structure in networks is fundamental for systems in biology, physics, and engineering. Community detection identifies candidate modules as hypotheses, which then need to be validated through experiments, such as mutagenesis in a biological laboratory. Only a few communities can typically be validated, and it is thus important to prioritize which communities to select for downstream experimentation. Here we develop CRANK, a mathematically principled approach for prioritizing network communities. CRANK efficiently evaluates robustness and magnitude of structural features of each community and then combines these features into the community prioritization. CRANK can be used with any community detection method. It needs only information provided by the network structure and does not require any additional metadata or labels. However, when available, CRANK can incorporate domain-specific information to further boost performance. Experiments on many large networks show that CRANK effectively prioritizes communities, yielding a nearly 50-fold improvement in community prioritization.

Genomewide phenotypic analysis of growth, cell morphogenesis, and cell cycle events in Escherichia coli

Cell size, cell growth, and cell cycle events are necessarily intertwined to achieve robust bacterial replication. Yet, a comprehensive and integrated view of these fundamental processes is lacking. Here, we describe an image‐based quantitative screen of the single‐gene knockout collection of Escherichia coli and identify many new genes involved in cell morphogenesis, population growth, nucleoid (bulk chromosome) dynamics, and cell division. Functional analyses, together with high‐dimensional classification, unveil new associations of morphological and cell cycle phenotypes with specific functions and pathways. Additionally, correlation analysis across ~4,000 genetic perturbations shows that growth rate is surprisingly not predictive of cell size. Growth rate was also uncorrelated with the relative timings of nucleoid separation and cell constriction. Rather, our analysis identifies scaling relationships between cell size and nucleoid size and between nucleoid size and the relative timings of nucleoid separation and cell division. These connections suggest that the nucleoid links cell morphogenesis to the cell cycle.

Features of the Chaperone Cellular Network Revealed through Systematic Interaction Mapping

A comprehensive view of molecular chaperone function in the cell was obtained through a systematic global integrative network approach based on physical (protein-protein) and genetic (gene-gene or epistatic) interaction mapping. This allowed us to decipher interactions involving all core chaperones (67) and cochaperones (15) of Saccharomyces cerevisiae. Our analysis revealed the presence of a large chaperone functional supercomplex, which we named the naturally joined (NAJ) chaperone complex, encompassing Hsp40, Hsp70, Hsp90, AAA+, CCT, and small Hsps. We further found that many chaperones interact with proteins that form foci or condensates under stress conditions. Using an in vitro reconstitution approach, we demonstrate condensate formation for the highly conserved AAA+ ATPases Rvb1 and Rvb2, which are part of the R2TP complex that interacts with Hsp90. This expanded view of the chaperone network in the cell clearly demonstrates the distinction between chaperones having broad versus narrow substrate specificities in protein homeostasis.

Systematic analysis of complex genetic interactions

To systematically explore complex genetic interactions, we constructed ~200,000 yeast triple mutants and scored negative trigenic interactions. We selected double-mutant query genes across a broad spectrum of biological processes, spanning a range of quantitative features of the global digenic interaction network and tested for a genetic interaction with a third mutation. Trigenic interactions often occurred among functionally related genes, and essential genes were hubs on the trigenic network. Despite their functional enrichment, trigenic interactions tended to link genes in distant bioprocesses and displayed a weaker magnitude than digenic interactions. We estimate that the global trigenic interaction network is ~100 times as large as the global digenic network, highlighting the potential for complex genetic interactions to affect the biology of inheritance, including the genotype-to-phenotype relationship.

Toward an integrated map of genetic interactions in cancer cells

Cancer genomes often harbor hundreds of molecular aberrations. Such genetic variants can be drivers or passengers of tumorigenesis and create vulnerabilities for potential therapeutic exploitation. To identify genotype-dependent vulnerabilities, forward genetic screens in different genetic backgrounds have been conducted. We devised MINGLE, a computational framework to integrate CRISPR/Cas9 screens originating from different libraries building on approaches pioneered for genetic network discovery in model organisms. We applied this method to integrate and analyze data from 85 CRISPR/Cas9 screens in human cancer cells combining functional data with information on genetic variants to explore more than 2.1 million gene-background relationships. In addition to known dependencies, we identified new genotype-specific vulnerabilities of cancer cells. Experimental validation of predicted vulnerabilities identified GANAB and PRKCSH as new positive regulators of Wnt/β-catenin signaling. By clustering genes with similar genetic interaction profiles, we drew the largest genetic network in cancer cells to date. Our scalable approach highlights how diverse genetic screens can be integrated to systematically build informative maps of genetic interactions in cancer, which can grow dynamically as more data are included.

High-Density Proximity Mapping Reveals the Subcellular Organization of mRNA-Associated Granules and Bodies

mRNA processing, transport, translation, and ultimately degradation involve a series of dedicated protein complexes that often assemble into large membraneless structures such as stress granules (SGs) and processing bodies (PBs). Here, systematic in vivo proximity-dependent biotinylation (BioID) analysis of 119 human proteins associated with different aspects of mRNA biology uncovers 7424 unique proximity interactions with 1,792 proteins. Classical bait-prey analysis reveals connections of hundreds of proteins to distinct mRNA-associated processes or complexes, including the splicing and transcriptional elongation machineries (protein phosphatase 4) and the CCR4-NOT deadenylase complex (CEP85, RNF219, and KIAA0355). Analysis of correlated patterns between endogenous preys uncovers the spatial organization of RNA regulatory structures and enables the definition of 144 core components of SGs and PBs. We report preexisting contacts between most core SG proteins under normal growth conditions and demonstrate that several core SG proteins (UBAP2L, CSDE1, and PRRC2C) are critical for the formation of microscopically visible SGs.

Unification of Protein Abundance Datasets Yields a Quantitative Saccharomyces cerevisiae Proteome

Protein activity is the ultimate arbiter of function in most cellular pathways, and protein concentration is fundamentally connected to protein action. While the proteome of yeast has been subjected to the most comprehensive analysis of any eukaryote, existing datasets are difficult to compare, and there is no consensus abundance value for each protein. We evaluated 21 quantitative analyses of the S. cerevisiae proteome, normalizing and converting all measurements of protein abundance into the intuitive measurement of absolute molecules per cell. We estimate the cellular abundance of 92% of the proteins in the yeast proteome and assess the variation in each abundance measurement. Using our protein abundance dataset, we find that a global response to diverse environmental stresses is not detected at the level of protein abundance, we find that protein tags have only a modest effect on protein abundance, and we identify proteins that are differentially regulated at the mRNA abundance, mRNA translation, and protein abundance levels.

Spatial Analysis of Functional Enrichment (SAFE) in Large Biological Networks

Spatial analysis of functional enrichment (SAFE) is a systematic quantitative approach for annotating large biological networks. SAFE detects network regions that are statistically overrepresented for functional groups or quantitative phenotypes of interest, and provides an intuitive visual representation of their relative positioning within the network. In doing so, SAFE determines which functions cocluster in a network, which parts of the network they are associated with and how they are potentially related to one another.Here, I provide a detailed stepwise description of how to perform a SAFE analysis. As an example, I use SAFE to annotate the genome-scale genetic interaction similarity network from Saccharomyces cerevisiae with Gene Ontology (GO) biological process terms. In addition, I show how integrating GO with chemical genomic data in SAFE can recapitulate known modes of action of chemical compounds and potentially identify novel drug mechanisms.

Computational Analysis of the Chaperone Interaction Networks

We provide computational protocols to identify chaperone interacting proteins using a combination of both physical (protein-protein) and genetic (gene-gene or epistatic) interaction data derived from the published large-scale proteomic and genomic studies for the budding yeast Saccharomyces cerevisiae. Using these datasets, we discuss bioinformatic analyses that can be employed to build comprehensive high-fidelity chaperone interaction networks. Given that many proteins typically function as complexes in the cell, we highlight various step-wise approaches for combining both the genetic and physical interaction datasets to decipher intra- and inter-connections for distinct chaperone- and non-chaperone-containing complexes in the network. Together, these informatics procedures will aid in identifying protein complexes with distinctive functional specializations in the cell that yield a very broad and diverse set of interactions. The described procedures can also be leveraged to datasets from other eukaryotes, including humans.

Functional annotation of chemical libraries across diverse biological processes

Chemical-genetic approaches offer the potential for unbiased functional annotation of chemical libraries. Mutations can alter the response of cells in the presence of a compound, revealing chemical-genetic interactions that can elucidate a compound's mode of action. We developed a highly parallel, unbiased yeast chemical-genetic screening system involving three key components. First, in a drug-sensitive genetic background, we constructed an optimized diagnostic mutant collection that is predictive for all major yeast biological processes. Second, we implemented a multiplexed (768-plex) barcode-sequencing protocol, enabling the assembly of thousands of chemical-genetic profiles. Finally, based on comparison of the chemical-genetic profiles with a compendium of genome-wide genetic interaction profiles, we predicted compound functionality. Applying this high-throughput approach, we screened seven different compound libraries and annotated their functional diversity. We further validated biological process predictions, prioritized a diverse set of compounds, and identified compounds that appear to have dual modes of action.

Protein-protein interfaces are vdW dominant with selective H-bonds and (or) electrostatics towards broad functional specificity

Several catalysis, cellular regulation, immune function, cell wall assembly, transport, signaling and inhibition occur through Protein- Protein Interactions (PPI). This is possible with the formation of specific yet stable protein-protein interfaces. Therefore, it is of interest to understand its molecular principles using structural data in relation to known function. Several interface features have been documented using known X-ray structures of protein complexes since 1975. This has improved our understanding of the interface using structural features such as interface area, binding energy, hydrophobicity, relative hydrophobicity, salt bridges and hydrogen bonds. The strength of binding between two proteins is dependent on interface size (number of residues at the interface) and thus its corresponding interface area. It is known that large interfaces have high binding energy (sum of (van der Waals) vdW, H-bonds, electrostatics). However, the selective role played by each of these energy components and more especially that of vdW is not explicitly known. Therefore, it is important to document their individual role in known protein-protein structural complexes. It is of interest to relate interface size with vdW, H-bonds and electrostatic interactions at the interfaces of protein structural complexes with known function using statistical and multiple linear regression analysis methods to identify the prominent force. We used the manually curated non-redundant dataset of 278 hetero-dimeric protein structural complexes grouped using known functions by Sowmya et al. (2015) to gain additional insight to this phenomenon using a robust inter-atomic non-covalent interaction analyzing tool PPCheck (Anshul and Sowdhamini, 2015). This dataset consists of obligatory (enzymes, regulator, biological assembly), immune and nonobligatory (enzyme and regulator inhibitors) complexes. Results show that the total binding energy is more for large interfaces. However, this is not true for its individual energy factors. Analysis shows that vdW energies contribute to about 75% ± 11% on average among all complexes and it also increases with interface size (r2 ranging from 0.67 to 0.89 with p<0.01) at 95% confidence limit irrespective of molecular function. Thus, vdW is both dominant and proportional at the interface independent of molecular function. Nevertheless, H bond energy contributes to 15% ± 6.5% on average in these complexes. It also moderately increases with interface size (r2 ranging from 0.43 to 0.61 with p<0.01) only among obligatory and immune complexes. Moreover, there is about 11.3% ± 8.7% contribution by electrostatic energy. It increases with interface size specifically among non-obligatory regulator-inhibitors (r2 = 0.44). It is implied that both H-bonds and electrostatics are neither dominant nor proportional at the interface. Nonetheless, their presence cannot be ignored in binding. Therefore, H-bonds and (or) electrostatic energy having specific role for improved stability in complexes is implied. Thus, vdW is common at the interface stabilized further with selective H-bonds and (or) electrostatic interactions at an atomic level in almost all complexes. Comparison of this observation with residue level analysis of the interface is compelling. The role by H-bonds (14.83% ± 6.5% and r2 = 0.61 with p<0.01) among obligatory and electrostatic energy (8.8% ± 4.77% and r2 = 0.63 with p <0.01) among non-obligatory complexes within interfaces (class A) having more non-polar residues than surface is influencing our inference. However, interfaces (class B) having less non-polar residues than surface show 1.5 fold more electrostatic energy on average. The interpretation of the interface using inter-atomic (vdW, H-bonds, electrostatic) interactions combined with inter-residue predominance (class A and class B) in relation to known function is the key to reveal its molecular principles with new challenges.

TheCellMap.org: A Web-Accessible Database for Visualizing and Mining the Global Yeast Genetic Interaction Network

Providing access to quantitative genomic data is key to ensure large-scale data validation and promote new discoveries. TheCellMap.org serves as a central repository for storing and analyzing quantitative genetic interaction data produced by genome-scale Synthetic Genetic Array (SGA) experiments with the budding yeast Saccharomyces cerevisiae. In particular, TheCellMap.org allows users to easily access, visualize, explore, and functionally annotate genetic interactions, or to extract and reorganize subnetworks, using data-driven network layouts in an intuitive and interactive manner.

PTMOracle: A Cytoscape App for Covisualizing and Coanalyzing Post-Translational Modifications in Protein Interaction Networks

Post-translational modifications of proteins (PTMs) act as key regulators of protein activity and of protein-protein interactions (PPIs). To date, it has been difficult to comprehensively explore functional links between PTMs and PPIs. To address this, we developed PTMOracle, a Cytoscape app for coanalyzing PTMs within PPI networks. PTMOracle also allows extensive data to be integrated and coanalyzed with PPI networks, allowing the role of domains, motifs, and disordered regions to be considered. For proteins of interest, or a whole proteome, PTMOracle can generate network visualizations to reveal complex PTM-associated relationships. This is assisted by OraclePainter for coloring proteins by modifications, OracleTools for network analytics, and OracleResults for exploring tabulated findings. To illustrate the use of PTMOracle, we investigate PTM-associated relationships and their role in PPIs in four case studies. In the yeast interactome and its rich set of PTMs, we construct and explore histone-associated and domain-domain interaction networks and show how integrative approaches can predict kinases involved in phosphodegrons. In the human interactome, a phosphotyrosine-associated network is analyzed but highlights the sparse nature of human PPI networks and lack of PTM-associated data. PTMOracle is open source and available at the Cytoscape app store: http://apps.cytoscape.org/apps/ptmoracle .

A global genetic interaction network maps a wiring diagram of cellular function

We generated a global genetic interaction network for Saccharomyces cerevisiae, constructing more than 23 million double mutants, identifying about 550,000 negative and about 350,000 positive genetic interactions. This comprehensive network maps genetic interactions for essential gene pairs, highlighting essential genes as densely connected hubs. Genetic interaction profiles enabled assembly of a hierarchical model of cell function, including modules corresponding to protein complexes and pathways, biological processes, and cellular compartments. Negative interactions connected functionally related genes, mapped core bioprocesses, and identified pleiotropic genes, whereas positive interactions often mapped general regulatory connections among gene pairs, rather than shared functionality. The global network illustrates how coherent sets of genetic interactions connect protein complex and pathway modules to map a functional wiring diagram of the cell.