Complexome profiling allows large-scale, untargeted, and comprehensive characterization of protein complexes in a biological sample using a combined approach of separating intact protein complexes e.g., by native gel electrophoresis, followed by mass spectrometric analysis of the proteins in the resulting fractions. Over the last decade, its application has resulted in a large collection of complexome profiling datasets. While computational methods have been developed for the analysis of individual datasets, methods for large-scale comparative analysis of complexomes from multiple species are lacking. Here, we present Comparative Clustering (CompaCt), that performs fully automated integrative analysis of complexome profiling data from multiple species, enabling systematic characterization and comparison of complexomes. CompaCt implements a novel method for leveraging orthology in comparative analysis to allow systematic identification of conserved as well as taxon-specific elements of the analyzed complexomes. We applied this method to a collection of 53 complexome profiles spanning the major branches of the eukaryotes. We demonstrate the ability of CompaCt to robustly identify the composition of protein complexes, and show that integrated analysis of multiple datasets improves characterization of complexes from specific complexome profiles when compared to separate analyses. We identified novel candidate interactors and complexes in a number of species from previously analyzed datasets, like the emp24, the V-ATPase and mitochondrial ATP synthase complexes. Lastly, we demonstrate the utility of CompaCt for the automated large-scale characterization of the complexome of the mosquito Anopheles stephensi shedding light on the evolution of metazoan protein complexes. CompaCt is available from https://github.com/cmbi/compact-bio.

Plasmodium species, the causative agent of malaria, have a complex life cycle involving two hosts. The sporozoite life stage is characterized by an extended phase in the mosquito salivary glands followed by free movement and rapid invasion of hepatocytes in the human host. This transmission stage has been the subject of many transcriptomics and proteomics studies and is also targeted by the most advanced malaria vaccine. We applied Bayesian data integration to determine which proteins are not only present in sporozoites but are also specific to that stage. Transcriptomic and proteomic Plasmodium data sets from 26 studies were weighted for how representative they are for sporozoites, based on a carefully assembled gold standard for Plasmodium falciparum (Pf) proteins known to be present or absent during the sporozoite life stage. Of 5418 Pf genes for which expression data were available at the RNA level or at the protein level, 975 were identified as enriched in sporozoites and 90 specific to them. We show that Pf sporozoites are enriched for proteins involved in type II fatty acid synthesis in the apicoplast and GPI anchor synthesis, but otherwise appear metabolically relatively inactive in the salivary glands of mosquitos. Newly annotated hypothetical sporozoite-specific and sporozoite-enriched proteins highlight sporozoite-specific functions. They include PF3D7_0104100 that we identified to be homologous to the prominin family, which in human has been related to a quiescent state of cancer cells. We document high levels of genetic variability for sporozoite proteins, specifically for sporozoite-specific proteins that elicit antibodies in the human host. Nevertheless, we can identify nine relatively well-conserved sporozoite proteins that elicit antibodies and that together can serve as markers for previous exposure. Our understanding of sporozoite biology benefits from identifying key pathways that are enriched during this life stage. This work can guide studies of molecular mechanisms underlying sporozoite biology and potential well-conserved targets for marker and drug development.

MOTIVATION: Complexome profiling combines native gel electrophoresis with mass spectrometry to obtain the inventory, composition and abundance of multiprotein assemblies in an organelle. Applying complexome profiling to determine the effect of a mutation on protein complexes requires separating technical and biological variations from the variations caused by that mutation. RESULTS: We have developed the COmplexome Profiling ALignment (COPAL) tool that aligns multiple complexome profiles with each other. It includes the abundance profiles of all proteins on two gels, using a multi-dimensional implementation of the dynamic time warping algorithm to align the gels. Subsequent progressive alignment allows us to align multiple profiles with each other. We tested COPAL on complexome profiles from control mitochondria and from Barth syndrome (BTHS) mitochondria, which have a mutation in tafazzin gene that is involved in remodeling the inner mitochondrial membrane phospholipid cardiolipin. By comparing the variation between BTHS mitochondria and controls with the variation among either, we assessed the effects of BTHS on the abundance profiles of individual proteins. Combining those profiles with gene set enrichment analysis allows detecting significantly affected protein complexes. Most of the significantly affected protein complexes are located in the inner mitochondrial membrane (mitochondrial contact site and cristae organizing system, prohibitins), or are attached to it (the large ribosomal subunit). AVAILABILITY AND IMPLEMENTATION: COPAL is written in python and is available from http://github.com/cmbi/copal. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Protein families participating in protein-protein interactions may contain sub-families that have different binding characteristics, ranging from right binding to showing no interaction at all. Composition differences at the sequence level in these sub-families are often decisive to their differential functional interaction. Methods to predict interface sites from protein sequences typically exploit conservation as a signal. Here, instead, we provide proof of concept that the sequence specificity between interacting versus non-interacting groups can be exploited to recognise interaction sites. RESULTS: We collected homodimeric and monomeric proteins and formed homologous groups, each having an interacting (homodimer) subgroup and a non-interacting (monomer) subgroup. We then compiled multiple sequence alignments of the proteins in the homologous groups and identified compositional differences between the homodimeric and monomeric subgroups for each of the alignment positions. Our results show that this specificity signal distinguishes interface and other surface residues with 40.9% recall and up to 25.1% precision. CONCLUSIONS: To our best knowledge, this is the first large scale study that exploits sequence specificity between interacting and non-interacting homologs to predict interaction sites from sequence information only. The performance obtained indicates that this signal contains valuable information to identify protein-protein interaction sites.

The RIG-I-like receptor (RLR) pathway is essential for detecting cytosolic viral RNA to trigger the production of type I interferons (IFNα/β) that initiate an innate antiviral response. Through systematic assessment of a wide variety of genomics data, we discovered 10 molecular signatures of known RLR pathway components that collectively predict novel members. We demonstrate that RLR pathway genes, among others, tend to evolve rapidly, interact with viral proteins, contain a limited set of protein domains, are regulated by specific transcription factors, and form a tightly connected interaction network. Using a Bayesian approach to integrate these signatures, we propose likely novel RLR regulators. RNAi knockdown experiments revealed a high prediction accuracy, identifying 94 genes among 187 candidates tested (~50%) that affected viral RNA-induced production of IFNβ. The discovered antiviral regulators may participate in a wide range of processes that highlight the complexity of antiviral defense (e.g. MAP3K11, CDK11B, PSMA3, TRIM14, HSPA9B, CDC37, NUP98, G3BP1), and include uncharacterized factors (DDX17, C6orf58, C16orf57, PKN2, SNW1). Our validated RLR pathway list (http://rlr.cmbi.umcn.nl/), obtained using a combination of integrative genomics and experiments, is a new resource for innate antiviral immunity research.

The large size of metabolic networks entails an overwhelming multiplicity in the possible steady-state flux distributions that are compatible with stoichiometric constraints. This space of possibilities is largest in the frequent situation where the nutrients available to the cells are unknown. These two factors: network size and lack of knowledge of nutrient availability, challenge the identification of the actual metabolic state of living cells among the myriad possibilities. Here we address this challenge by developing a method that integrates gene-expression measurements with genome-scale models of metabolism as a means of inferring metabolic states. Our method explores the space of alternative flux distributions that maximize the agreement between gene expression and metabolic fluxes, and thereby identifies reactions that are likely to be active in the culture from which the gene-expression measurements were taken. These active reactions are used to build environment-specific metabolic models and to predict actual metabolic states. We applied our method to model the metabolic states of Saccharomyces cerevisiae growing in rich media supplemented with either glucose or ethanol as the main energy source. The resulting models comprise about 50% of the reactions in the original model, and predict environment-specific essential genes with high sensitivity. By minimizing the sum of fluxes while forcing our predicted active reactions to carry flux, we predicted the metabolic states of these yeast cultures that are in large agreement with what is known about yeast physiology. Most notably, our method predicts the Crabtree effect in yeast cells growing in excess glucose, a long-known phenomenon that could not have been predicted by traditional constraint-based modeling approaches. Our method is of immediate practical relevance for medical and industrial applications, such as the identification of novel drug targets, and the development of biotechnological processes that use complex, largely uncharacterized media, such as biofuel production.

Gene duplication is one of the main mechanisms by which genomes can acquire novel functions. It has been proposed that the retention of gene duplicates can be associated to processes of tissue expression divergence. These models predict that acquisition of divergent expression patterns should be acquired shortly after the duplication, and that larger divergence in tissue expression would be expected for paralogs, as compared to orthologs of a similar age. Many studies have shown that gene duplicates tend to have divergent expression patterns and that gene family expansions are associated with high levels of tissue specificity. However, the timeframe in which these processes occur have rarely been investigated in detail, particularly in vertebrates, and most analyses do not include direct comparisons of orthologs as a baseline for the expected levels of tissue specificity in absence of duplications. To assess the specific contribution of duplications to expression divergence, we combine here phylogenetic analyses and expression data from human and mouse. In particular, we study differences in spatial expression among human-mouse paralogs, specifically duplicated after the radiation of mammals, and compare them to pairs of orthologs in the same species. Our results show that gene duplication leads to increased levels of tissue specificity and that this tends to occur promptly after the duplication event.

MOTIVATION: The intensification of DNA sequencing will increasingly unveil uncharacterized species with potential alternative genetic codes. A total of 0.65% of the DNA sequences currently in Genbank encode their proteins with a variant genetic code, and these exceptions occur in many unrelated taxa. RESULTS: We introduce FACIL (Fast and Accurate genetic Code Inference and Logo), a fast and reliable tool to evaluate nucleic acid sequences for their genetic code that detects alternative codes even in species distantly related to known organisms. To illustrate this, we apply FACIL to a set of mitochondrial genomic contigs of Globobulimina pseudospinescens. This foraminifer does not have any sequenced close relative in the databases, yet we infer its alternative genetic code with high confidence values. Results are intuitively visualized in a Genetic Code Logo. AVAILABILITY AND IMPLEMENTATION: FACIL is available as a web-based service at http://www.cmbi.ru.nl/FACIL/ and as a stand-alone program.

MOTIVATION: Protein-protein interaction (PPI) networks are a valuable resource for the interpretation of genomics data. However, such networks have interaction enrichment biases for proteins that are often studied. These biases skew quantitative results from comparing PPI networks with genomics data. Here, we introduce an approach named physical interaction enrichment (PIE) to eliminate these biases. METHODOLOGY: PIE employs a normalization that ensures equal node degree (edge) distribution of a test set and of the random networks it is compared with. It quantifies whether a set of proteins have more interactions between themselves than proteins in random networks, and can therewith be regarded as physically cohesive. RESULTS: Among other datasets, we applied PIE to genetic morbid disease (GMD) genes and to genes whose expression is induced upon infection with human-metapneumovirus (HMPV). Both sets contain proteins that are often studied and that have relatively many interactions in the PPI network. Although interactions between proteins of both sets are found to be overrepresented in PPI networks, the GMD proteins are not more likely to interact with each other than random proteins when this overrepresentation is taken into account. In contrast the HMPV-induced genes, representing a biologically more coherent set, encode proteins that do tend to interact with each other and can be used to predict new HMPV-induced genes. By handling biases in PPI networks, PIE can be a valuable tool to quantify the degree to which a set of genes are involved in the same biological process.

MOTIVATION: Most microbial species can not be cultured in the laboratory. Metagenomic sequencing may still yield a complete genome if the sequenced community is enriched and the sequencing coverage is high. However, the complexity in a natural population may cause the enrichment culture to contain multiple related strains. This diversity can confound existing strict assembly programs and lead to a fragmented assembly, which is unnecessary if we have a related reference genome available that can function as a scaffold. RESULTS: Here, we map short metagenomic sequencing reads from a population of strains to a related reference genome, and compose a genome that captures the consensus of the population's sequences. We show that by iteration of the mapping and assembly procedure, the coverage increases while the similarity with the reference genome decreases. This indicates that the assembly becomes less dependent on the reference genome and approaches the consensus genome of the multi-strain population. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Genes that are co-expressed tend to be involved in the same biological process. However, co-expression is not a very reliable predictor of functional links between genes. The evolutionary conservation of co-expression between species can be used to predict protein function more reliably than co-expression in a single species. Here we examine whether co-expression across multiple species is also a better prioritizer of disease genes than is co-expression between human genes alone. RESULTS: We use co-expression data from yeast (S. cerevisiae), nematode worm (C. elegans), fruit fly (D. melanogaster), mouse and human and find that the use of evolutionary conservation can indeed improve the predictive value of co-expression. The effect that genes causing the same disease have higher co-expression than do other genes from their associated disease loci, is significantly enhanced when co-expression data are combined across evolutionarily distant species. We also find that performance can vary significantly depending on the co-expression datasets used, and just using more data does not necessarily lead to better prioritization. Instead, we find that dataset quality is more important than quantity, and using a consistent microarray platform per species leads to better performance than using more inclusive datasets pooled from various platforms. CONCLUSION: We find that evolutionarily conserved gene co-expression prioritizes disease candidate genes better than human gene co-expression alone, and provide the integrated data as a new resource for disease gene prioritization tools.

MOTIVATION: Phylogenomics integrates the vast amount of phylogenetic information contained in complete genome sequences, and is rapidly becoming the standard for reliably inferring species phylogenies. There are, however, fundamental differences between the ways in which phylogenomic approaches like gene content, superalignment, superdistance and supertree integrate the phylogenetic information from separate orthologous groups. Furthermore, they all depend on the method by which the orthologous groups are initially determined. Here, we systematically compare these four phylogenomic approaches, in parallel with three approaches for large-scale orthology determination: pairwise orthology, cluster orthology and tree-based orthology. RESULTS: Including various phylogenetic methods, we apply a total of 54 fully automated phylogenomic procedures to the fungi, the eukaryotic clade with the largest number of sequenced genomes, for which we retrieved a golden standard phylogeny from the literature. Phylogenomic trees based on gene content show, relative to the other methods, a bias in the tree topology that parallels convergence in lifestyle among the species compared, indicating convergence in gene content. CONCLUSIONS: Complete genomes are no guarantee for good or even consistent phylogenies. However, the large amounts of data in genomes enable us to carefully select the data most suitable for phylogenomic inference. In terms of performance, the superalignment approach, combined with restrictive orthology, is the most successful in recovering a fungal phylogeny that agrees with current taxonomic views, and allows us to obtain a high-resolution phylogeny. We provide solid support for what has grown to be a common practice in phylogenomics during its advance in recent years. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Orthology is one of the cornerstones of gene function prediction. Dividing the phylogenetic relations between genes into either orthologs or paralogs is however an oversimplification. Already in two-species gene-phylogenies, the complicated, non-transitive nature of phylogenetic relations results in inparalogs and outparalogs. For situations with more than two species we lack semantics to specifically describe the phylogenetic relations, let alone to exploit them. Published procedures to extract orthologous groups from phylogenetic trees do not allow identification of orthology at various levels of resolution, nor do they document the relations between the orthologous groups. RESULTS: We introduce "levels of orthology" to describe the multi-level nature of gene relations. This is implemented in a program LOFT (Levels of Orthology From Trees) that assigns hierarchical orthology numbers to genes based on a phylogenetic tree. To decide upon speciation and gene duplication events in a tree LOFT can be instructed either to perform classical species-tree reconciliation or to use the species overlap between partitions in the tree. The hierarchical orthology numbers assigned by LOFT effectively summarize the phylogenetic relations between genes. The resulting high-resolution orthologous groups are depicted in colour, facilitating visual inspection of (large) trees. A benchmark for orthology prediction, that takes into account the varying levels of orthology between genes, shows that the phylogeny-based high-resolution orthology assignments made by LOFT are reliable. CONCLUSION: The "levels of orthology" concept offers high resolution, reliable orthology, while preserving the relations between orthologous groups. A Windows as well as a preliminary Java version of LOFT is available from the LOFT website http://www.cmbi.ru.nl/LOFT.

Mitochondria are eukaryotic organelles that originated from the endosymbiosis of an alpha-proteobacterium. To gain insight into the evolution of the mitochondrial proteome as it proceeded through the transition from a free-living cell to a specialized organelle, we compared a reconstructed ancestral proteome of the mitochondrion with the proteomes of alpha-proteobacteria as well as with the mitochondrial proteomes in yeast and man. Overall, there has been a large turnover of the mitochondrial proteome during the evolution of mitochondria. Early in the evolution of the mitochondrion, proteins involved in cell envelope synthesis have virtually disappeared, whereas proteins involved in replication, transcription, cell division, transport, regulation, and signal transduction have been replaced by eukaryotic proteins. More than half of what remains from the mitochondrial ancestor in modern mitochondria corresponds to translation, including post-translational modifications, and to metabolic pathways that are directly, or indirectly, involved in energy conversion. Altogether, the results indicate that the eukaryotic host has hijacked the proto-mitochondrion, taking control of its protein synthesis and metabolism.

MOTIVATION: Databases of protein families often exhibit drastically different properties of the protein family space. RESULTS: We compared the properties of protein family space as reflected by exhaustive protein family databases and databases with predefined families. We used TRIBES, Protomap, ProDom and COGs as representatives of the exhaustive databases, and Pfam-A and Superfamily as databases that predefine families. We observe a power-law distribution of family sizes in all these databases, albeit in predefined databases the power-law line collapses before reaching smaller sized families. We discuss the future trends of this power-law distribution and suggest that saturation in the sampling of protein family space will result in a distortion of the power law in small family sizes. For larger genome sizes, predefined databases show logarithmic growth of the number of families per genome, whereas exhaustive databases exhibit a virtually linear relationship. All databases consistently differ in the proportion of protein families shared between taxa. Predefined databases have a larger number of protein families shared between the three domains of life, while exhaustive databases show a much more fragmented distribution. We argue that these discrepancies reflect alternative approaches to the trade-off issue of sensitivity versus specificity in the detection of homologous proteins. We conclude that these properties are complementary rather than contradictory, while describing the protein universe from different perspectives.

Abstract Motivation: Several genes involved in the cellular import of copper and its subsequent incorporation into the high-affinity iron transport complex in Saccharomyces cerevisiae are known to be conserved between eukaryotes and prokaryotes. However, the degree to which these genes share their functional context as members of the same pathway in the prokaryotic domain is less clear. Results: The co-occurrence of gene families involved in Atx1p-mediated copper transport in the genomes and operon structures of 80 non-redundant prokaryotes was investigated. For this purpose, we developed a Web tool (SHOPS) to display the operon context for a given set of proteins. In total, a set of 43 putative operons was identified. These were found to be involved in a variety of pathways and indicate a large diversity in the functional context of the individual gene family members. Availability: The SHOPS tool can be found at http://www.bioinformatics.med.uu.nl/shops. Supplemental data are available at http://humgen.med.uu.nl/publications/vanbakel/pathway/ Supplementary information: For additional information on our datasets and for the supplemental figures referred to in the text, please see http://humgen.med.uu.nl/publications/vanbakel/pathway/. The Web-based tool, SHOPS, can be found at http://www.bioinformatics.med.uu.nl/shops/