Explore the vast range of titles available.

Background Predicting protein function and structure from sequence is one important challenge for computational biology. For 26 years, most state-of-the-art approaches combined machine learning and evolutionary information. However, for some applications retrieving related proteins is becoming too time-consuming. Additionally, evolutionary information is less powerful for small families, e.g. for proteins from the Dark Proteome . Both these problems are addressed by the new methodology introduced here. Results We introduced a novel way to represent protein sequences as continuous vectors ( embeddings ) by using the language model ELMo taken from natural language processing. By modeling protein sequences, ELMo effectively captured the biophysical properties of the language of life from unlabeled big data (UniRef50). We refer to these new embeddings as SeqVec ( Seq uence-to- Vec tor) and demonstrate their effectiveness by training simple neural networks for two different tasks. At the per-residue level, secondary structure (Q3 = 79% ± 1, Q8 = 68% ± 1) and regions with intrinsic disorder (MCC = 0.59 ± 0.03) were predicted significantly better than through one-hot encoding or through Word2vec-like approaches. At the per-protein level, subcellular localization was predicted in ten classes (Q10 = 68% ± 1) and membrane-bound were distinguished from water-soluble proteins (Q2 = 87% ± 1). Although SeqVec embeddings generated the best predictions from single sequences, no solution improved over the best existing method using evolutionary information. Nevertheless, our approach improved over some popular methods using evolutionary information and for some proteins even did beat the best. Thus, they prove to condense the underlying principles of protein sequences. Overall, the important novelty is speed: where the lightning-fast HHblits needed on average about two minutes to generate the evolutionary information for a target protein, SeqVec created embeddings on average in 0.03 s. As this speed-up is independent of the size of growing sequence databases, SeqVec provides a highly scalable approach for the analysis of big data in proteomics, i.e. microbiome or metaproteome analysis. Conclusion Transfer-learning succeeded to extract information from unlabeled sequence databases relevant for various protein prediction tasks. SeqVec modeled the language of life, namely the principles underlying protein sequences better than any features suggested by textbooks and prediction methods. The exception is evolutionary information, however, that information is not available on the level of a single sequence.

Culture-independent analyses of microbial communities have progressed dramatically in the last decade, particularly due to advances in methods for biological profiling via shotgun metagenomics. Opportunities for improvement continue to accelerate, with greater access to multi-omics, microbial reference genomes, and strain-level diversity. To leverage these, we present bioBakery 3, a set of integrated, improved methods for taxonomic, strain-level, functional, and phylogenetic profiling of metagenomes newly developed to build on the largest set of reference sequences now available. Compared to current alternatives, MetaPhlAn 3 increases the accuracy of taxonomic profiling, and HUMAnN 3 improves that of functional potential and activity. These methods detected novel disease-microbiome links in applications to CRC (1262 metagenomes) and IBD (1635 metagenomes and 817 metatranscriptomes). Strain-level profiling of an additional 4077 metagenomes with StrainPhlAn 3 and PanPhlAn 3 unraveled the phylogenetic and functional structure of the common gut microbe Ruminococcus bromii , previously described by only 15 isolate genomes. With open-source implementations and cloud-deployable reproducible workflows, the bioBakery 3 platform can help researchers deepen the resolution, scale, and accuracy of multi-omic profiling for microbial community studies.

Protein phase separation is implicated in formation of membraneless organelles, signaling puncta and the nuclear pore. Multivalent interactions of modular binding domains and their target motifs can drive phase separation. However, forces promoting the more common phase separation of intrinsically disordered regions are less understood, with suggested roles for multivalent cation-pi, pi-pi, and charge interactions and the hydrophobic effect. Known phase-separating proteins are enriched in pi-orbital containing residues and thus we analyzed pi-interactions in folded proteins. We found that pi-pi interactions involving non-aromatic groups are widespread, underestimated by force-fields used in structure calculations and correlated with solvation and lack of regular secondary structure, properties associated with disordered regions. We present a phase separation predictive algorithm based on pi interaction frequency, highlighting proteins involved in biomaterials and RNA processing.

Recent developments in deep learning, coupled with an increasing number of sequenced proteins, have led to a breakthrough in life science applications, in particular in protein property prediction. There is hope that deep learning can close the gap between the number of sequenced proteins and proteins with known properties based on lab experiments. Language models from the field of natural language processing have gained popularity for protein property predictions and have led to a new computational revolution in biology, where old prediction results are being improved regularly. Such models can learn useful multipurpose representations of proteins from large open repositories of protein sequences and can be used, for instance, to predict protein properties. The field of natural language processing is growing quickly because of developments in a class of models based on a particular model—the Transformer model. We review recent developments and the use of large-scale Transformer models in applications for predicting protein characteristics and how such models can be used to predict, for example, post-translational modifications. We review shortcomings of other deep learning models and explain how the Transformer models have quickly proven to be a very promising way to unravel information hidden in the sequences of amino acids.

Predicting the function of a protein from its amino acid sequence is a long-standing challenge in bioinformatics. Traditional approaches use sequence alignment to compare a query sequence either to thousands of models of protein families or to large databases of individual protein sequences. Here we introduce ProteInfer, which instead employs deep convolutional neural networks to directly predict a variety of protein functions – Enzyme Commission (EC) numbers and Gene Ontology (GO) terms – directly from an unaligned amino acid sequence. This approach provides precise predictions which complement alignment-based methods, and the computational efficiency of a single neural network permits novel and lightweight software interfaces, which we demonstrate with an in-browser graphical interface for protein function prediction in which all computation is performed on the user’s personal computer with no data uploaded to remote servers. Moreover, these models place full-length amino acid sequences into a generalised functional space, facilitating downstream analysis and interpretation. To read the interactive version of this paper, please visit https://google-research.github.io/proteinfer/ .

Predicting the thermodynamic stability of proteins is a common and widely used step in protein engineering, and when elucidating the molecular mechanisms behind evolution and disease. Here, we present RaSP, a method for making rapid and accurate predictions of changes in protein stability by leveraging deep learning representations. RaSP performs on-par with biophysics-based methods and enables saturation mutagenesis stability predictions in less than a second per residue. We use RaSP to calculate ∼ 230 million stability changes for nearly all single amino acid changes in the human proteome, and examine variants observed in the human population. We find that variants that are common in the population are substantially depleted for severe destabilization, and that there are substantial differences between benign and pathogenic variants, highlighting the role of protein stability in genetic diseases. RaSP is freely available—including via a Web interface—and enables large-scale analyses of stability in experimental and predicted protein structures.

Contemporary bioinformatic and chemoinformatic capabilities hold promise to reshape knowledge management, analysis and interpretation of data in natural products research. Currently, reliance on a disparate set of non-standardized, insular, and specialized databases presents a series of challenges for data access, both within the discipline and for integration and interoperability between related fields. The fundamental elements of exchange are referenced structure-organism pairs that establish relationships between distinct molecular structures and the living organisms from which they were identified. Consolidating and sharing such information via an open platform has strong transformative potential for natural products research and beyond. This is the ultimate goal of the newly established LOTUS initiative, which has now completed the first steps toward the harmonization, curation, validation and open dissemination of 750,000+ referenced structure-organism pairs. LOTUS data is hosted on Wikidata and regularly mirrored on https://lotus.naturalproducts.net . Data sharing within the Wikidata framework broadens data access and interoperability, opening new possibilities for community curation and evolving publication models. Furthermore, embedding LOTUS data into the vast Wikidata knowledge graph will facilitate new biological and chemical insights. The LOTUS initiative represents an important advancement in the design and deployment of a comprehensive and collaborative natural products knowledge base.

MicroRNA targets are often recognized through pairing between the miRNA seed region and complementary sites within target mRNAs, but not all of these canonical sites are equally effective, and both computational and in vivo UV-crosslinking approaches suggest that many mRNAs are targeted through non-canonical interactions. Here, we show that recently reported non-canonical sites do not mediate repression despite binding the miRNA, which indicates that the vast majority of functional sites are canonical. Accordingly, we developed an improved quantitative model of canonical targeting, using a compendium of experimental datasets that we pre-processed to minimize confounding biases. This model, which considers site type and another 14 features to predict the most effectively targeted mRNAs, performed significantly better than existing models and was as informative as the best high-throughput in vivo crosslinking approaches. It drives the latest version of TargetScan (v7.0; targetscan.org), thereby providing a valuable resource for placing miRNAs into gene-regulatory networks. Proteins are built by using the information contained in molecules of messenger RNA (mRNA). Cells have several ways of controlling the amounts of different proteins they make. For example, a so-called ‘microRNA’ molecule can bind to an mRNA molecule to cause it to be more rapidly degraded and less efficiently used, thereby reducing the amount of protein built from that mRNA. Indeed, microRNAs are thought to help control the amount of protein made from most human genes, and biologists are working to predict the amount of control imparted by each microRNA on each of its mRNA targets. All RNA molecules are made up of a sequence of bases, each commonly known by a single letter—‘A’, ‘U’, ‘C’ or ‘G’. These bases can each pair up with one specific other base—‘A’ pairs with ‘U’, and ‘C’ pairs with ‘G’. To direct the repression of an mRNA molecule, a region of the microRNA known as a ‘seed’ binds to a complementary sequence in the target mRNA. ‘Canonical sites’ are regions in the mRNA that contain the exact sequence of partner bases for the bases in the microRNA seed. Some canonical sites are more effective at mRNA control than others. ‘Non-canonical sites’ also exist in which the pairing between the microRNA seed and mRNA does not completely match. Previous work has suggested that many non-canonical sites can also control mRNA degradation and usage. Agarwal et al. first used large experimental datasets from many sources to investigate microRNA activity in more detail. As expected, when mRNAs had canonical sites that matched the microRNA, mRNA levels and usage tended to drop. However, no effect was observed when the mRNAs only had recently identified non-canonical sites. This suggests that microRNAs primarily bind to canonical sites to control protein production. Based on these results, Agarwal et al. further developed a statistical model that predicts the effects of microRNAs binding to canonical sites. The updated model considers 14 different features of the microRNA, microRNA site, or mRNA—including the mRNA sequence around the site—to predict which sites within mRNAs are most effectively targeted by microRNAs. Tests showed that Agarwal et al.'s model was as good as experimental approaches at identifying the effective target sites, and was better than existing computational models. The model has been used to power the latest version of a freely available resource called TargetScan, and so could prove a valuable resource for researchers investigating the many important roles of microRNAs in controlling protein production.

Replicability is an important feature of scientific research, but aspects of contemporary research culture, such as an emphasis on novelty, can make replicability seem less important than it should be. The Reproducibility Project: Cancer Biology was set up to provide evidence about the replicability of preclinical research in cancer biology by repeating selected experiments from high-impact papers. A total of 50 experiments from 23 papers were repeated, generating data about the replicability of a total of 158 effects. Most of the original effects were positive effects (136), with the rest being null effects (22). A majority of the original effect sizes were reported as numerical values (117), with the rest being reported as representative images (41). We employed seven methods to assess replicability, and some of these methods were not suitable for all the effects in our sample. One method compared effect sizes: for positive effects, the median effect size in the replications was 85% smaller than the median effect size in the original experiments, and 92% of replication effect sizes were smaller than the original. The other methods were binary – the replication was either a success or a failure – and five of these methods could be used to assess both positive and null effects when effect sizes were reported as numerical values. For positive effects, 40% of replications (39/97) succeeded according to three or more of these five methods, and for null effects 80% of replications (12/15) were successful on this basis; combining positive and null effects, the success rate was 46% (51/112). A successful replication does not definitively confirm an original finding or its theoretical interpretation. Equally, a failure to replicate does not disconfirm a finding, but it does suggest that additional investigation is needed to establish its reliability.