Achieving high accuracy in orthology inference is essential for many comparative, evolutionary and functional genomic analyses, yet the true evolutionary history of genes is generally unknown and orthologs are used for very different applications across phyla, requiring different precision-recall trade-offs. As a result, it is difficult to assess the performance of orthology inference methods. Here, we present a community effort to establish standards and an automated web-based service to facilitate orthology benchmarking. Using this service, we characterize 15 well-established inference methods and resources on a battery of 20 different benchmarks. Standardized benchmarking provides a way for users to identify the most effective methods for the problem at hand, sets a minimum requirement for new tools and resources, and guides the development of more accurate orthology inference methods.

The mentioning of gene names in the body of the scientific literature 1901-2017 and their fractional counting is used as a proxy to assess the level of biological function discovery. A literature score of one has been defined as full publication equivalent (FPE), the amount of literature necessary to achieve one publication solely dedicated to a gene. It has been found that less than 5000 human genes have each at least 100 FPEs in the available literature corpus. This group of elite genes (4817 protein-coding genes, 119 non-coding RNAs) attracts the overwhelming majority of the scientific literature about genes. Yet, thousands of proteins have never been mentioned at all, ≈2000 further proteins have not even one FPE of literature and, for ≈4600 additional proteins, the FPE count is below 10. The protein function discovery rate measured as numbers of proteins first mentioned or crossing a threshold of accumulated FPEs in a given year has grown until 2000 but is in decline thereafter. This drop is partially offset by function discoveries for non-coding RNAs. The full human genome sequencing does not boost the function discovery rate. Since 2000, the fastest growing group in the literature is that with at least 500 FPEs per gene.

Circular RNAs (circRNAs) are increasingly recognized to play crucial roles in post-transcriptional gene regulation including functioning as microRNA (miRNA) sponges or as wide-spread regulators, for example in stem cell differentiation. It is therefore highly relevant to identify if a transcript of interest can also function as a circRNA. Here, we present a user-friendly web server that predicts if coding and noncoding RNAs have circRNA isoforms and whether circRNAs are expressed in stem cells. The predictions are made by random forest models using sequence-derived features as input. The output scores are converted to fractiles, which are used to assess the circRNA and stem cell potential. The performances of the three models are reported as the area under the receiver operating characteristic (ROC) curve and are 0.82 for coding genes, 0.89 for long noncoding RNAs (lncRNAs) and 0.72 for stem cell expression. We present WebCircRNA for quick evaluation of human genes and transcripts for their circRNA potential, which can be essential in several contexts.

Most BioCreative tasks to date have focused on assessing the quality of text-mining annotations in terms of precision and recall. Interoperability, speed, and stability are, however, other important factors to consider for practical applications of text mining. For about a decade, we have run named entity recognition (NER) web services, which are designed to be efficient, implemented using a multi-threaded queueing system to robustly handle many simultaneous requests, and hosted at a supercomputer facility. To participate in this new task, we extended the existing NER tagging service with support for the BeCalm API. The tagger suffered no downtime during the challenge and, as in earlier tests, proved to be highly efficient, consistently processing requests of 5000 abstracts in less than half a minute. In fact, the majority of this time was spent not on the NER task but rather on retrieving the document texts from the challenge servers. The latter was found to be the main bottleneck even when hosting a copy of the tagging service on a Raspberry Pi 3, showing that local document storage or caching would be desirable features to include in future revisions of the API standard.

Many protein domains bind to short peptide sequences, called linear motifs. Data on their sequence specificities is sparse, which is why biologists usually resort to basic pattern searches to identify new putative binding sites for experimental follow-up. Most motifs have poor specificity and prioritization of the matches is thus crucial when scanning a full proteome with a pattern. Here we present a generic method to prioritize motif occurrence predictions by using cellular contextual information. We take 2 parameters as input: the motif occurrences and one or more of the interacting domains. The potential hits are ranked based on how strongly the context network associates them with a protein containing one of the specified domains, which leads to an increased predictive performance. The method is available through a web interface at doremi.jensenlab.org, which allows for an easy application of the method. We show that this approach leads to improved predictions of binding partners for PDZ domains and the SUMO binding domain. This is consistent with the earlier observation that coupling sequence motifs with network information improves kinase-specific substrate predictions.

The 'druggable genome' encompasses several protein families, but only a subset of targets within them have attracted significant research attention and thus have information about them publicly available. The Illuminating the Druggable Genome (IDG) program was initiated in 2014, has the goal of developing experimental techniques and a Knowledge Management Center (KMC) that would collect and organize information about protein targets from four families, representing the most common druggable targets with an emphasis on understudied proteins. Here, we describe two resources developed by the KMC: the Target Central Resource Database (TCRD) which collates many heterogeneous gene/protein datasets and Pharos (https://pharos.nih.gov), a multimodal web interface that presents the data from TCRD. We briefly describe the types and sources of data considered by the KMC and then highlight features of the Pharos interface designed to enable intuitive access to the IDG knowledgebase. The aim of Pharos is to encourage 'serendipitous browsing', whereby related, relevant information is made easily discoverable. We conclude by describing two use cases that highlight the utility of Pharos and TCRD.

A great challenge in the proteomics and structural genomics era is to predict protein structure and function, including identification of those proteins that are partially or wholly unstructured. Disordered regions in proteins often contain short linear peptide motifs (e.g., SH3 ligands and targeting signals) that are important for protein function. We present here DisEMBL, a computational tool for prediction of disordered/unstructured regions within a protein sequence. As no clear definition of disorder exists, we have developed parameters based on several alternative definitions and introduced a new one based on the concept of "hot loops," i.e., coils with high temperature factors. Avoiding potentially disordered segments in protein expression constructs can increase expression, foldability, and stability of the expressed protein. DisEMBL is thus useful for target selection and the design of constructs as needed for many biochemical studies, particularly structural biology and structural genomics projects. The tool is freely available via a web interface (http://dis.embl.de) and can be downloaded for use in large-scale studies.

Understanding the distribution of taxa and associated traits across different environments is one of the central questions in microbial ecology. High-throughput sequencing (HTS) studies are presently generating huge volumes of data to address this biogeographical topic. However, these studies are often focused on specific environment types or processes leading to the production of individual, unconnected datasets. The large amounts of legacy sequence data with associated metadata that exist can be harnessed to better place the genetic information found in these surveys into a wider environmental context. Here we introduce a software program, seqenv, to carry out precisely such a task. It automatically performs similarity searches of short sequences against the "nt" nucleotide database provided by NCBI and, out of every hit, extracts-if it is available-the textual metadata field. After collecting all the isolation sources from all the search results, we run a text mining algorithm to identify and parse words that are associated with the Environmental Ontology (EnvO) controlled vocabulary. This, in turn, enables us to determine both in which environments individual sequences or taxa have previously been observed and, by weighted summation of those results, to summarize complete samples. We present two demonstrative applications of seqenv to a survey of ammonia oxidizing archaea as well as to a plankton paleome dataset from the Black Sea. These demonstrate the ability of the tool to reveal novel patterns in HTS and its utility in the fields of environmental source tracking, paleontology, and studies of microbial biogeography. To install seqenv, go to: https://github.com/xapple/seqenv.

Diabetes is a diverse and complex disease, with considerable variation in phenotypic manifestation and severity. This variation hampers the study of etiological differences and reduces the statistical power of analyses of associations to genetics, treatment outcomes, and complications. We address these issues through deep, fine-grained phenotypic stratification of a diabetes cohort. Text mining the electronic health records of 14,017 patients, we matched two controlled vocabularies (ICD-10 and a custom vocabulary developed at the clinical center Steno Diabetes Center Copenhagen) to clinical narratives spanning a 19 year period. The two matched vocabularies comprise over 20,000 medical terms describing symptoms, other diagnoses, and lifestyle factors. The cohort is genetically homogeneous (Caucasian diabetes patients from Denmark) so the resulting stratification is not driven by ethnic differences, but rather by inherently dissimilar progression patterns and lifestyle related risk factors. Using unsupervised Markov clustering, we defined 71 clusters of at least 50 individuals within the diabetes spectrum. The clusters display both distinct and shared longitudinal glycemic dysregulation patterns, temporal co-occurrences of comorbidities, and associations to single nucleotide polymorphisms in or near genes relevant for diabetes comorbidities.

Although Escherichia coli (E. coli) is the most studied prokaryote organism in the history of life sciences, many molecular mechanisms and gene functions encoded in its genome remain to be discovered. This work aims at quantifying the illumination of the E. coli gene function space by the scientific literature and how close we are towards the goal of a complete list of E. coli gene functions.

eggNOG is a public resource that provides Orthologous Groups (OGs) of proteins at different taxonomic levels, each with integrated and summarized functional annotations. Developments since the latest public release include changes to the algorithm for creating OGs across taxonomic levels, making nested groups hierarchically consistent. This allows for a better propagation of functional terms across nested OGs and led to the novel annotation of 95 890 previously uncharacterized OGs, increasing overall annotation coverage from 67% to 72%. The functional annotations of OGs have been expanded to also provide Gene Ontology terms, KEGG pathways and SMART/Pfam domains for each group. Moreover, eggNOG now provides pairwise orthology relationships within OGs based on analysis of phylogenetic trees. We have also incorporated a framework for quickly mapping novel sequences to OGs based on precomputed HMM profiles. Finally, eggNOG version 4.5 incorporates a novel data set spanning 2605 viral OGs, covering 5228 proteins from 352 viral proteomes. All data are accessible for bulk downloading, as a web-service, and through a completely redesigned web interface. The new access points provide faster searches and a number of new browsing and visualization capabilities, facilitating the needs of both experts and less experienced users. eggNOG v4.5 is available at http://eggnog.embl.de.

Orthology assignment is ideally suited for functional inference. However, because predicting orthology is computationally intensive at large scale, and most pipelines are relatively inaccessible (e.g., new assignments only available through database updates), less precise homology-based functional transfer is still the default for (meta-)genome annotation. We, therefore, developed eggNOG-mapper, a tool for functional annotation of large sets of sequences based on fast orthology assignments using precomputed clusters and phylogenies from the eggNOG database. To validate our method, we benchmarked Gene Ontology (GO) predictions against two widely used homology-based approaches: BLAST and InterProScan. Orthology filters applied to BLAST results reduced the rate of false positive assignments by 11%, and increased the ratio of experimentally validated terms recovered over all terms assigned per protein by 15%. Compared with InterProScan, eggNOG-mapper achieved similar proteome coverage and precision while predicting, on average, 41 more terms per protein and increasing the rate of experimentally validated terms recovered over total term assignments per protein by 35%. EggNOG-mapper predictions scored within the top-5 methods in the three GO categories using the CAFA2 NK-partial benchmark. Finally, we evaluated eggNOG-mapper for functional annotation of metagenomics data, yielding better performance than interProScan. eggNOG-mapper runs ∼15× faster than BLAST and at least 2.5× faster than InterProScan. The tool is available standalone and as an online service at http://eggnog-mapper.embl.de.

The subcellular localization of a protein is an important aspect of its function. However, the experimental annotation of locations is not even complete for well-studied model organisms. Text mining might aid database curators to add experimental annotations from the scientific literature. Existing extraction methods have difficulties to distinguish relationships between proteins and cellular locations co-mentioned in the same sentence.

The number of discovered natural miRNA sponges in plants, viruses, and mammals is increasing steadily. Some sponges like ciRS-7 for miR-7 contain multiple nearby miRNA binding sites. We hypothesize that such clusters of miRNA binding sites on the genome can function together as a sponge. No systematic effort has been made in search for clusters of miRNA targets. Here, we, to our knowledge, make the first genome-wide target site predictions for clusters of mature human miRNAs. For each miRNA, we predict the target sites on a genome-wide scale, build a graph with edge weights based on the pairwise distances between sites, and apply Markov clustering to identify genomic regions with high binding site density. Significant clusters are then extracted based on cluster size difference between real and shuffled genomes preserving local properties such as the GC content. We then use conservation and binding energy to filter a final set of miRNA target site clusters or sponge candidates. Our pipeline predicts 3673 sponge candidates for 1250 miRNAs, including the experimentally verified miR-7 sponge ciRS-7. In addition, we point explicitly to 19 high-confidence candidates overlapping annotated genomic sequence. The full list of candidates is freely available at http://rth.dk/resources/mirnasponge, where detailed properties for individual candidates can be explored, such as alignment details, conservation, accessibility and target profiles, which facilitates selection of sponge candidates for further context specific analysis.

To facilitate the study of interactions between proteins and chemicals, we have created STITCH, an aggregated database of interactions connecting over 300,000 chemicals and 2.6 million proteins from 1133 organisms. Compared to the previous version, the number of chemicals with interactions and the number of high-confidence interactions both increase 4-fold. The database can be accessed interactively through a web interface, displaying interactions in an integrated network view. It is also available for computational studies through downloadable files and an API. As an extension in the current version, we offer the option to switch between two levels of detail, namely whether stereoisomers of a given compound are shown as a merged entity or as separate entities. Separate display of stereoisomers is necessary, for example, for carbohydrates and chiral drugs. Combining the isomers increases the coverage, as interaction databases and publications found through text mining will often refer to compounds without specifying the stereoisomer. The database is accessible at http://stitch.embl.de/.

For most sequenced prokaryotic genomes, about a third of the protein coding genes annotated are "orphan proteins", that is, they lack homology to known proteins. These hypothetical genes are typically short and randomly scattered throughout the genome. This trend is seen for most of the bacterial and archaeal genomes published to date.

Over the last years, the publicly available knowledge on interactions between small molecules and proteins has been steadily increasing. To create a network of interactions, STITCH aims to integrate the data dispersed over the literature and various databases of biological pathways, drug-target relationships and binding affinities. In STITCH 2, the number of relevant interactions is increased by incorporation of BindingDB, PharmGKB and the Comparative Toxicogenomics Database. The resulting network can be explored interactively or used as the basis for large-scale analyses. To facilitate links to other chemical databases, we adopt InChIKeys that allow identification of chemicals with a short, checksum-like string. STITCH 2.0 connects proteins from 630 organisms to over 74,000 different chemicals, including 2200 drugs. STITCH can be accessed at http://stitch.embl.de/.

Protein kinases control cellular decision processes by phosphorylating specific substrates. Thousands of in vivo phosphorylation sites have been identified, mostly by proteome-wide mapping. However, systematically matching these sites to specific kinases is presently infeasible, due to limited specificity of consensus motifs, and the influence of contextual factors, such as protein scaffolds, localization, and expression, on cellular substrate specificity. We have developed an approach (NetworKIN) that augments motif-based predictions with the network context of kinases and phosphoproteins. The latter provides 60%-80% of the computational capability to assign in vivo substrate specificity. NetworKIN pinpoints kinases responsible for specific phosphorylations and yields a 2.5-fold improvement in the accuracy with which phosphorylation networks can be constructed. Applying this approach to DNA damage signaling, we show that 53BP1 and Rad50 are phosphorylated by CDK1 and ATM, respectively. We describe a scalable strategy to evaluate predictions, which suggests that BCLAF1 is a GSK-3 substrate.

Implementing precision medicine hinges on the integration of omics data, such as proteomics, into the clinical decision-making process, but the quantity and diversity of biomedical data, and the spread of clinically relevant knowledge across multiple biomedical databases and publications, pose a challenge to data integration. Here we present the Clinical Knowledge Graph (CKG), an open-source platform currently comprising close to 20 million nodes and 220 million relationships that represent relevant experimental data, public databases and literature. The graph structure provides a flexible data model that is easily extendable to new nodes and relationships as new databases become available. The CKG incorporates statistical and machine learning algorithms that accelerate the analysis and interpretation of typical proteomics workflows. Using a set of proof-of-concept biomarker studies, we show how the CKG might augment and enrich proteomics data and help inform clinical decision-making.

Protein association networks can be inferred from a range of resources including experimental data, literature mining and computational predictions. These types of evidence are emerging for non-coding RNAs (ncRNAs) as well. However, integration of ncRNAs into protein association networks is challenging due to data heterogeneity. Here, we present a database of ncRNA-RNA and ncRNA-protein interactions and its integration with the STRING database of protein-protein interactions. These ncRNA associations cover four organisms and have been established from curated examples, experimental data, interaction predictions and automatic literature mining. RAIN uses an integrative scoring scheme to assign a confidence score to each interaction. We demonstrate that RAIN outperforms the underlying microRNA-target predictions in inferring ncRNA interactions. RAIN can be operated through an easily accessible web interface and all interaction data can be downloaded.Database URL: http://rth.dk/resources/rain.