MOTIVATION: Representation learning has revolutionized sequence-based prediction of protein function and subcellular localization. Protein networks are an important source of information complementary to sequences, but the use of protein networks has proven to be challenging in the context of machine learning, especially in a cross-species setting. RESULTS: We leveraged the STRING database of protein networks and orthology relations for 1322 eukaryotes to generate network-based cross-species protein embeddings. We did this by first creating species-specific network embeddings and subsequently aligning them based on orthology relations to facilitate direct cross-species comparisons. We show that these aligned network embeddings ensure consistency across species without sacrificing quality compared to species-specific network embeddings. We also show that the aligned network embeddings are complementary to sequence embedding techniques, despite the use of sequence-based orthology relations in the alignment process. Finally, we validated the embeddings by using them for two well-established tasks: subcellular localization prediction and protein function prediction. Training logistic regression classifiers on aligned network embeddings and sequence embeddings improved the accuracy over using sequence alone, reaching performance numbers close to state-of-the-art deep-learning methods. AVAILABILITY AND IMPLEMENTATION: The source code and scripts for generating the network-based cross-species protein embeddings are available at https://github.com/deweihu96/SPACE. Precomputed network embeddings and sequence embeddings for all eukaryotic proteins are included in STRING version 12.0 (https://string-db.org/cgi/download).

A knowledge-based grouping of genes into pathways or functional units is essential for describing and understanding cellular complexity. However, it is not always clear a priori how and at what level of specificity functionally interconnected genes should be partitioned into pathways, for a given application. Here, we assess and compare nine existing and two conceptually novel functional classification systems, with respect to their discovery power and generality in gene set enrichment testing. We base our assessment on a collection of nearly 2000 functional genomics datasets provided by users of the STRING database. With these real-life and diverse queries, we assess which systems typically provide the most specific and complete enrichment results. We find many structural and performance differences between classification systems. Overall, the well-established, hierarchically organized pathway annotation systems yield the best enrichment performance, despite covering substantial parts of the human genome in general terms only. On the other hand, the more recent unsupervised annotation systems perform strongest in understudied areas and organisms, and in detecting more specific pathways, albeit with less informative labels.

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: An orthologous group (OG) comprises a set of orthologous and paralogous genes that share a last common ancestor (LCA). OGs are defined with respect to a chosen taxonomic level, which delimits the position of the LCA in time to a specified speciation event. A hierarchy of OGs expands on this notion, connecting more general OGs, distant in time, to more recent, fine-grained OGs, thereby spanning multiple levels of the tree of life. Large scale inference of OG hierarchies with independently computed taxonomic levels can suffer from inconsistencies between successive levels, such as the position in time of a duplication event. This can be due to confounding genetic signal or algorithmic limitations. Importantly, inconsistencies limit the potential use of OGs for functional annotation and third-party applications. RESULTS: Here we present a new methodology to ensure hierarchical consistency of OGs across taxonomic levels. To resolve an inconsistency, we subsample the protein space of the OG members and perform gene tree-species tree reconciliation for each sampling. Differently from previous approaches, by subsampling the protein space, we avoid the notoriously difficult task of accurately building and reconciling very large phylogenies. We implement the method into a high-throughput pipeline and apply it to the eggNOG database. We use independent protein domain definitions to validate its performance. CONCLUSION: The presented consistency pipeline shows that, contrary to previous limitations, tree reconciliation can be a useful instrument for the construction of OG hierarchies. The key lies in the combination of sampling smaller trees and aggregating their reconciliations for robustness. Results show comparable or greater performance to previous pipelines. The code is available on Github at: https://github.com/meringlab/og_consistency_pipeline .

MOTIVATION: Ribosomal RNA profiling has become crucial to studying microbial communities, but meaningful taxonomic analysis and inter-comparison of such data are still hampered by technical limitations, between-study design variability and inconsistencies between taxonomies used. RESULTS: Here we present MAPseq, a framework for reference-based rRNA sequence analysis that is up to 30% more accurate (F½ score) and up to one hundred times faster than existing solutions, providing in a single run multiple taxonomy classifications and hierarchical operational taxonomic unit mappings, for rRNA sequences in both amplicon and shotgun sequencing strategies, and for datasets of virtually any size. AVAILABILITY AND IMPLEMENTATION: Source code and binaries are freely available at https://github.com/jfmrod/mapseq. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

UNLABELLED: A successful approach for predicting functional associations between non-homologous genes is to compare their phylogenetic distributions. We have devised a phylogenetic profiling algorithm, SVD-Phy, which uses truncated singular value decomposition to address the problem of uninformative profiles giving rise to false positive predictions. Benchmarking the algorithm against the KEGG pathway database, we found that it has substantially improved performance over existing phylogenetic profiling methods. AVAILABILITY AND IMPLEMENTATION: The software is available under the open-source BSD license at https://bitbucket.org/andrea/svd-phy CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Nucleotide sequence data are being produced at an ever increasing rate. Clustering such sequences by similarity is often an essential first step in their analysis-intended to reduce redundancy, define gene families or suggest taxonomic units. Exact clustering algorithms, such as hierarchical clustering, scale relatively poorly in terms of run time and memory usage, yet they are desirable because heuristic shortcuts taken during clustering might have unintended consequences in later analysis steps. RESULTS: Here we present HPC-CLUST, a highly optimized software pipeline that can cluster large numbers of pre-aligned DNA sequences by running on distributed computing hardware. It allocates both memory and computing resources efficiently, and can process more than a million sequences in a few hours on a small cluster. AVAILABILITY AND IMPLEMENTATION: Source code and binaries are freely available at http://meringlab.org/software/hpc-clust/; the pipeline is implemented in Cþþ and uses the Message Passing Interface (MPI) standard for distributed computing.

Operational Taxonomic Units (OTUs), usually defined as clusters of similar 16S/18S rRNA sequences, are the most widely used basic diversity units in large-scale characterizations of microbial communities. However, it remains unclear how well the various proposed OTU clustering algorithms approximate 'true' microbial taxa. Here, we explore the ecological consistency of OTUs--based on the assumption that, like true microbial taxa, they should show measurable habitat preferences (niche conservatism). In a global and comprehensive survey of available microbial sequence data, we systematically parse sequence annotations to obtain broad ecological descriptions of sampling sites. Based on these, we observe that sequence-based microbial OTUs generally show high levels of ecological consistency. However, different OTU clustering methods result in marked differences in the strength of this signal. Assuming that ecological consistency can serve as an objective external benchmark for cluster quality, we conclude that hierarchical complete linkage clustering, which provided the most ecologically consistent partitions, should be the default choice for OTU clustering. To our knowledge, this is the first approach to assess cluster quality using an external, biologically meaningful parameter as a benchmark, on a global scale.

Non-intermingling, adjacent populations of cells define compartment boundaries; such boundaries are often essential for the positioning and the maintenance of tissue-organizers during growth. In the developing wing primordium of Drosophila melanogaster, signaling by the secreted protein Hedgehog (Hh) is required for compartment boundary maintenance. However, the precise mechanism of Hh input remains poorly understood. Here, we combine experimental observations of perturbed Hh signaling with computer simulations of cellular behavior, and connect physical properties of cells to their Hh signaling status. We find that experimental disruption of Hh signaling has observable effects on cell sorting surprisingly far from the compartment boundary, which is in contrast to a previous model that confines Hh influence to the compartment boundary itself. We have recapitulated our experimental observations by simulations of Hh diffusion and transduction coupled to mechanical tension along cell-to-cell contact surfaces. Intriguingly, the best results were obtained under the assumption that Hh signaling cannot alter the overall tension force of the cell, but will merely re-distribute it locally inside the cell, relative to the signaling status of neighboring cells. Our results suggest a scenario in which homotypic interactions of a putative Hh target molecule at the cell surface are converted into a mechanical force. Such a scenario could explain why the mechanical output of Hh signaling appears to be confined to the compartment boundary, despite the longer range of the Hh molecule itself. Our study is the first to couple a cellular vertex model describing mechanical properties of cells in a growing tissue, to an explicit model of an entire signaling pathway, including a freely diffusible component. We discuss potential applications and challenges of such an approach.

SUMMARY: The Helmholtz Network for Bioinformatics (HNB) is a joint venture of eleven German bioinformatics research groups that offers convenient access to numerous bioinformatics resources through a single web portal. The 'Guided Solution Finder' which is available through the HNB portal helps users to locate the appropriate resources to answer their queries by employing a detailed, tree-like questionnaire. Furthermore, automated complex tool cascades ('tasks'), involving resources located on different servers, have been implemented, allowing users to perform comprehensive data analyses without the requirement of further manual intervention for data transfer and re-formatting. Currently, automated cascades for the analysis of regulatory DNA segments as well as for the prediction of protein functional properties are provided. AVAILABILITY: The HNB portal is available at http://www.hnbioinfo.de

MOTIVATION: The analysis of protein-protein interactions allows for detailed exploration of the cellular machinery. The biochemical purification of protein complexes followed by identification of components by mass spectrometry is currently the method, which delivers the most reliable information--albeit that the data sets are still difficult to interpret. Consolidating individual experiments into protein complexes, especially for high-throughput screens, is complicated by many contaminants, the occurrence of proteins in otherwise dissimilar purifications due to functional re-use and technical limitations in the detection. A non-redundant collection of protein complexes from experimental data would be useful for biological interpretation, but manual assembly is tedious and often inconsistent. RESULTS: Here, we introduce a measure to define similarity within collections of purifications and generate a set of minimally redundant, comprehensive complexes using unsupervised clustering. AVAILABILITY: Programs and results are freely available from http://www.bork.embl-heidelberg.de/Docu/purclust/