Novel machine learning approaches revolutionize protein knowledge

Breakthrough methods in machine learning (ML), protein structure prediction, and novel ultrafast structural aligners are revolutionizing structural biology. Obtaining accurate models of proteins and annotating their functions on a large scale is no longer limited by time and resources. The most recent method to be top ranked by the Critical Assessment of Structure Prediction (CASP) assessment, AlphaFold 2 (AF2), is capable of building structural models with an accuracy comparable to that of experimental structures. Annotations of 3D models are keeping pace with the deposition of the structures due to advancements in protein language models (pLMs) and structural aligners that help validate these transferred annotations. In this review we describe how recent developments in ML for protein science are making large-scale structural bioinformatics available to the general scientific community.     

Nucleotide aggregation in aqueous solution. A multicomponent self-diffusion study

The self-aggregation of the mononucleotides (AMP, CMP, GMP and UMP) and caffeine up to their solubility limit in 2H2O has been monitored through self-diffusion measurements, using the Fourier transform NMR pulsed-gradient spin-echo self-diffusion technique. The data were iteratively fitted to a number of aggregation models. It was concluded that the best agreement between simulations and experiment for the mononucleotides was obtained for a 'semi-isodesmic', indefinite aggregation model (also known as a Type III SEK or cooperative indefinite self-association model), where the first (dimerization) aggregation constant is a magnitude lower than those for the higher aggregation steps. Typical values were 0.4 and 6 l mol-1, respectively. Under these conditions, the main fraction of solute is monomeric throughout the concentration range and the distribution of higher oligomers is very broad. Caffeine self-aggregation is clearly different and is consistent with several aggregation models. The mixed aggregation of caffeine (at a low total concentration) and the mononucleotides was successfully monitored in an extension of the basic study. It was found that caffeine binding to mononucleotide aggregates increases in the series UMP, CMP, GMP and AMP.     

The method matters: A guide for indicator aggregation in ecological assessments

Ecological assessment requires the integration of many physical, chemical, and/or biological quality elements. The choice of the aggregation method of such partial assessments into an overall assessment can considerably affect the assessment outcome – an issue that has been controversially discussed within the scientific community for the last decade. Current practice often considers only two different aggregation methods, the weighted arithmetic mean (additive aggregation) and the one-out, all-out method (minimum aggregation). However, both have important drawbacks. Additive aggregation compensates a bad status of one quality element by a number of elements featuring good status. Minimum aggregation can lead to overly pessimistic assessment results, since only the quality element in the worst status is considered. Here, we introduce a toolbox containing current and new aggregation methods, demonstrate and discuss their properties with simple, didactical examples, and suggest in which situations best to use them. Then, we illustrate the consequences of selected aggregation schemes for ecological river assessment with the case study of the Swiss Modular Concept of stream assessment (SMC), which we apply to ten river reaches in the Mönchaltdorfer Aa catchment in Switzerland. To be able to do so, we used multi-criteria decision analysis, i.e., multi-attribute value theory, to arrange the SMC quality elements into an objectives hierarchy, and to translate their individual assessments into value functions. Our case study revealed that choosing the most appropriate aggregation method particularly matters, if objectives with significantly different qualities are aggregated. We argue that redundant objectives (i.e., quality elements), often found at the lower levels of the objectives hierarchy, should best be aggregated additively allowing for compensation to increase the statistical significance of the results. Further, we suggest that complementary sub-objectives that often occur at higher levels may be optimally aggregated with a mixture of additive and minimum aggregation. Such a mixed method will allow some compensation, but nevertheless penalize for very bad states. Since here we compare commonly used aggregation methods with some which we believe have never been discussed in an assessment context before, our study concurrently informs ecological assessment in theory and in practice.     

PK/DB: database for pharmacokinetic properties and predictive in silico ADME models

The study of pharmacokinetic properties (PK) is of great importance in drug discovery and development. In the present work, PK/DB (a new freely available database for PK) was designed with the aim of creating robust databases for pharmacokinetic studies and in silico absorption, distribution, metabolism and excretion (ADME) prediction. Comprehensive, web-based and easy to access, PK/DB manages 1203 compounds which represent 2973 pharmacokinetic measurements, including five models for in silico ADME prediction (human intestinal absorption, human oral bioavailability, plasma protein binding, blood-brain barrier and water solubility). AVAILABILITY: http://www.pkdb.ifsc.usp.br     

CPAD,Curated Protein Aggregation Database: A Repository of Manually Curated Experimental Data on Protein and Peptide Aggregation

Accurate distinction between peptide sequences that can form amyloid-fibrils or amorphous β-aggregates, identification of potential aggregation prone regions in proteins, and prediction of change in aggregation rate of a protein upon mutation(s) are critical to research on protein misfolding diseases, such as Alzheimer’s and Parkinson’s, as well as biotechnological production of protein based therapeutics. We have developed a Curated Protein Aggregation Database (CPAD), which has collected results from experimental studies performed by scientific community aimed at understanding protein/peptide aggregation. CPAD contains more than 2300 experimentally observed aggregation rates upon mutations in known amyloidogenic proteins. Each entry includes numerical values for the following parameters: change in rate of aggregation as measured by fluorescence intensity or turbidity, name and source of the protein, Uniprot and Protein Data Bank codes, single point as well as multiple mutations, and literature citation. The data in CPAD has been supplemented with five different types of additional information: (i) Amyloid fibril forming hexa-peptides, (ii) Amorphous β-aggregating hexa-peptides, (iii) Amyloid fibril forming peptides of different lengths, (iv) Amyloid fibril forming hexa-peptides whose crystal structures are available in the Protein Data Bank (PDB) and (v) Experimentally validated aggregation prone regions found in amyloidogenic proteins. Furthermore, CPAD is linked to other related databases and resources, such as Uniprot, Protein Data Bank, PUBMED, GAP, TANGO, WALTZ etc. We have set up a web interface with different search and display options so that users have the ability to get the data in multiple ways. CPAD is freely available at http://www.iitm.ac.in/bioinfo/CPAD/. The potential applications of CPAD have also been discussed.     

Higher species diversity explained by stronger spatial aggregation across six neotropical Drosophila communities

Spatial aggregation of competitors over resource patches is generally accepted as an important mechanism maintaining coexistence of species in insect communities exploiting fragmented resources. However, its quantitative effects on local diversity, i.e. the relationship between the degree of aggregation in a community and community diversity, remain unexplored. In this paper, we tested whether stronger spatial aggregation does lead to the predicted higher local diversity. We compared six species-rich Drosophila communities exploiting decaying fruits in central Panama, monitored over one full year (> 25 generations). We found a clear positive relationship between the overall degree of aggregation and community diversity. In addition, aggregation over fruit trees was found to contribute greatly to the overall degree of aggregation and was largely responsible for the observed relationship between aggregation and diversity across communities. In addition, both diversity and aggregation strength were lower in communities in disturbed habitats, which was explained by altered spatial distribution of fruiting trees. This study shows that the aggregation model cannot only explain coexistence, but also differences in local diversity.     

Mfold web server for nucleic acid folding and hybridization prediction

The abbreviated name, 'mfold web server', describes a number of closely related software applications available on the World Wide Web (WWW) for the prediction of the secondary structure of single stranded nucleic acids. The objective of this web server is to provide easy access to RNA and DNA folding and hybridization software to the scientific community at large. By making use of universally available web GUIs (Graphical User Interfaces), the server circumvents the problem of portability of this software. Detailed output, in the form of structure plots with or without reliability information, single strand frequency plots and 'energy dot plots', are available for the folding of single sequences. A variety of 'bulk' servers give less information, but in a shorter time and for up to hundreds of sequences at once. The portal for the mfold web server is http://www.bioinfo.rpi.edu/applications/mfold. This URL will be referred to as 'MFOLDROOT'.     

Mining high-throughput experimental data to link gene and function

Nearly 2200 genomes that encode around 6 million proteins have now been sequenced. Around 40% of these proteins are of unknown function, even when function is loosely and minimally defined as 'belonging to a superfamily'. In addition to in silico methods, the swelling stream of high-throughput experimental data can give valuable clues for linking these unknowns with precise biological roles. The goal is to develop integrative data-mining platforms that allow the scientific community at large to access and utilize this rich source of experimental knowledge. To this end, we review recent advances in generating whole-genome experimental datasets, where this data can be accessed, and how it can be used to drive prediction of gene function.     

GAM模型和BRT模型在长江口鱼类群落多样性预测中的比较

长江口为西太平洋最大的河口,评估其鱼类群落多样性分布能够为长江口生态系统的修复和管理提供科学依据.本研究基于2012—2014年长江口渔业监测数据,分别使用GAM模型和BRT模型建立各站点水域鱼类群落多样性指数与环境和时空因子之间的关系.结合线性回归方程,采用交叉验证的方式对模型的预测能力和拟合效果进行评价,并绘制了2014年长江口鱼类群落多样性指数和丰富度指数的空间分布图.结果表明: 盐度、pH和叶绿素a对多样性指数贡献最高,pH、溶解氧和叶绿素a是对丰富度指数贡献率最高的环境因子.BRT模型对于多样性指数和丰富度指数的拟合和预测结果均优于GAM模型.空间分布预测显示,相较于GAM模型,BRT模型能够对长江口小面积水域间的鱼类群落多样性作更好的区分,河口外侧水域的鱼类群落多样性明显高于河口内侧水域,而北支水域的多样性高于南支水域.     

Identifying protein aggregation mechanisms and quantifying aggregation rates from combined monomer depletion and continuous scattering

Parallel temperature initial rates (PTIR) from chromatographic separation of aggregating protein solutions are combined with continuous simultaneous multiple sample light scattering (SMSLS) to make quantitative deductions about protein aggregation kinetics and mechanisms. PTIR determines the rates at which initially monomeric proteins are converted to aggregates over a range of temperatures, under initial-rate conditions. Using SMSLS for the same set of conditions provides time courses of the absolute Rayleigh scattering ratio, I_R(t), from which a potentially different measure of aggregation rates can be quantified. The present report compares these measures of aggregation rates across a range of solution conditions that result in different aggregation mechanisms for anti-streptavidin (AS) immunoglobulin gamma-1 (IgG1). The results illustrate how the two methods provide complementary information when deducing aggregation mechanisms, as well as cases where they provide new mechanistic details that were not possible to deduce in previous work. Criteria are presented for when the two techniques are expected to give equivalent results for quantitative rates, the potential limitations when solution non-idealities are large, as well as a comparison of the temperature dependence of AS-IgG1 aggregation rates with published data for other antibodies.     

The development of locomotor kinematics in neonatal rats: an agent-based modeling analysis in group and individual contexts

An agent-based model of infant rat (pup) locomotion and aggregation was developed by modifying a previous model of pup aggregation [Schank, J.C., Alberts, J.R., 2000a. The developmental emergence of coupled activity as cooperative aggregation in rat pups. Proc. R. Soc. London B 267, 2307-2315]. The main difference between the earlier and current models is the incorporation of whole-body kinematics of directional locomotion. Data on locomotion and aggregation are presented for individuals and groups of 7- and 10-day-old pups and the data were used to evolve models (with a genetic algorithm) that fit these data. Aggregation between 7- and 10-day-old pups was considerably different and could be explained by agent-based models, in particular, models with directional-kinematic matrices specifying the probabilities of moving to adjacent cells. The directional kinematics of whole-body movement differed between the two age classes and differed between group and individual contexts for 10-day-old pups. This may indicate a developmental transition (by day 10) to more central control of behavior and the ability to change patterns of movement based on social context. The behavior analyzed with agent-based models may provide a precise way to measure motor and nervous system development in rats and other rodents.     

Integrated decision-tree testing strategies for acute systemic toxicity and toxicokinetics with respect to the requirements of the EU REACH legislation

Liverpool John Moores University and FRAME conducted a joint research project, sponsored by Defra, on the status of alternatives to animal testing with regard to the European Union REACH (Registration, Evaluation and Authorisation of Chemicals) system for the safety testing and risk assessment of chemicals. The project covered all the main toxicity endpoints associated with REACH. This paper focuses on the use of alternative (non-animal) methods (both in vitro and in silico) for acute systemic toxicity and toxicokinetic testing. The paper reviews in vitro tests based on basal cytotoxicity and target organ toxicity, along with QSAR models and expert systems available for this endpoint. The use of PBPK modelling for the prediction of ADME properties is also discussed. These tests are then incorporated into a decision-tree style, integrated testing strategy, which also includes the use of refined in vivo acute toxicity tests, as a last resort. The implementation of the strategy is intended to minimise the use of animals in the testing of acute systemic toxicity and toxicokinetics, whilst satisfying the scientific and logistical demands of the EU REACH legislation.     

Transmembrane topology and signal peptide prediction using dynamic bayesian networks

Hidden Markov models (HMMs) have been successfully applied to the tasks of transmembrane protein topology prediction and signal peptide prediction. In this paper we expand upon this work by making use of the more powerful class of dynamic Bayesian networks (DBNs). Our model, Philius, is inspired by a previously published HMM, Phobius, and combines a signal peptide submodel with a transmembrane submodel. We introduce a two-stage DBN decoder that combines the power of posterior decoding with the grammar constraints of Viterbi-style decoding. Philius also provides protein type, segment, and topology confidence metrics to aid in the interpretation of the predictions. We report a relative improvement of 13% over Phobius in full-topology prediction accuracy on transmembrane proteins, and a sensitivity and specificity of 0.96 in detecting signal peptides. We also show that our confidence metrics correlate well with the observed precision. In addition, we have made predictions on all 6.3 million proteins in the Yeast Resource Center (YRC) database. This large-scale study provides an overall picture of the relative numbers of proteins that include a signal-peptide and/or one or more transmembrane segments as well as a valuable resource for the scientific community. All DBNs are implemented using the Graphical Models Toolkit. Source code for the models described here is available at http://noble.gs.washington.edu/proj/philius. A Philius Web server is available at http://www.yeastrc.org/philius, and the predictions on the YRC database are available at http://www.yeastrc.org/pdr.     

A new test set for validating predictions of protein-ligand interaction

We present a large test set of protein-ligand complexes for the purpose of validating algorithms that rely on the prediction of protein-ligand interactions. The set consists of 305 complexes with protonation states assigned by manual inspection. The following checks have been carried out to identify unsuitable entries in this set: (1) assessing the involvement of crystallographically related protein units in ligand binding; (2) identification of bad clashes between protein side chains and ligand; and (3) assessment of structural errors, and/or inconsistency of ligand placement with crystal structure electron density. In addition, the set has been pruned to assure diversity in terms of protein-ligand structures, and subsets are supplied for different protein-structure resolution ranges. A classification of the set by protein type is available. As an illustration, validation results are shown for GOLD and SuperStar. GOLD is a program that performs flexible protein-ligand docking, and SuperStar is used for the prediction of favorable interaction sites in proteins. The new CCDC/Astex test set is freely available to the scientific community (http://www.ccdc.cam.ac.uk).     

A unified model explains commonness and rarity on coral reefs

Abundance patterns in ecological communities have important implications for biodiversity maintenance and ecosystem functioning. However, ecological theory has been largely unsuccessful at capturing multiple macroecological abundance patterns simultaneously. Here, we propose a parsimonious model that unifies widespread ecological relationships involving local aggregation, species‐abundance distributions, and species associations, and we test this model against the metacommunity structure of reef‐building corals and coral reef fishes across the western and central Pacific. For both corals and fishes, the unified model simultaneously captures extremely well local species‐abundance distributions, interspecific variation in the strength of spatial aggregation, patterns of community similarity, species accumulation, and regional species richness, performing far better than alternative models also examined here and in previous work on coral reefs. Our approach contributes to the development of synthetic theory for large‐scale patterns of community structure in nature, and to addressing ongoing challenges in biodiversity conservation at macroecological scales.     

Evaluation of the taxonomic accuracy and pathogenicity prediction power of 16 primer sets amplifying single copy marker genes in the Pseudomonas syringae species complex

The Pseudomonas syringae species complex is composed of several closely related species of bacterial plant pathogens. Here, we used in silico methods to assess 16 PCR primer sets designed for broad identification of isolates throughout the species complex. We evaluated their in silico amplification rate in 2161 publicly available genomes, the correlation between pairwise amplicon sequence distance and whole genome average nucleotide identity, and trained naive Bayes classification models to quantify classification resolution. Furthermore, we show the potential for using single amplicon sequence data to predict type III effector protein repertoires, which are important determinants of host specificity and range.     

Dynamic imaging by fluorescence correlation spectroscopy identifies diverse populations of polyglutamine oligomers formed in vivo

Protein misfolding and aggregation are exacerbated by aging and diseases of protein conformation including neurodegeneration, metabolic diseases, and cancer. In the cellular environment, aggregates can exist as discrete entities, or heterogeneous complexes of diverse solubility and conformational state. In this study, we have examined the in vivo dynamics of aggregation using imaging methods including fluorescence microscopy, fluorescence recovery after photobleaching (FRAP), and fluorescence correlation spectroscopy (FCS), to monitor the diverse biophysical states of expanded polyglutamine (polyQ) proteins expressed in Caenorhabditis elegans. We show that monomers, oligomers and aggregates co-exist at different concentrations in young and aged animals expressing different polyQ-lengths. During aging, when aggregation and toxicity are exacerbated, FCS-based burst analysis and purified single molecule FCS detected a populational shift toward an increase in the frequency of brighter and larger oligomeric species. Regardless of age or polyQ-length, oligomers were maintained in a heterogeneous distribution that spans multiple orders of magnitude in brightness. We employed genetic suppressors that prevent polyQ aggregation and observed a reduction in visible immobile species with the persistence of heterogeneous oligomers, yet our analysis did not detect the appearance of any discrete oligomeric states associated with toxicity. These studies reveal that the reversible transition from monomers to immobile aggregates is not represented by discrete oligomeric states, but rather suggests that the process of aggregation involves a more complex pattern of molecular interactions of diverse intermediate species that can appear in vivo and contribute to aggregate formation and toxicity.     

Across species‐pool aggregation alters grassland productivity and diversity

Plant performance is determined by the balance of intra‐ and interspecific neighbors within an individual's zone of influence. If individuals interact over smaller scales than the scales at which communities are measured, then altering neighborhood interactions may fundamentally affect community responses. These interactions can be altered by changing the number (species richness), abundances (species evenness), and positions (species pattern) of the resident plant species, and we aimed to test whether aggregating species at planting would alter effects of species richness and evenness on biomass production at a common scale of observation in grasslands. We varied plant species richness (2, 4, or 8 species and monocultures), evenness (0.64, 0.8, or 1.0), and pattern (planted randomly or aggregated in groups of four individuals) within 1 × 1 m plots established with transplants from a pool of 16 tallgrass prairie species and assessed plot‐scale biomass production and diversity over the first three growing seasons. As expected, more species‐rich plots produced more biomass by the end of the third growing season, an effect associated with a shift from selection to complementarity effects over time. Aggregating conspecifics at a 0.25‐m scale marginally reduced biomass production across all treatments and increased diversity in the most even plots, but did not alter biodiversity effects or richness–productivity relationships. Results support the hypothesis that fine‐scale species aggregation affects diversity by promoting species coexistence in this system. However, results indicate that inherent changes in species neighborhood relationships along grassland diversity gradients may only minimally affect community (meter) – scale responses among similarly designed biodiversity–ecosystem function studies. Given that species varied in their responses to local aggregation, it may be possible to use such species‐specific results to spatially design larger‐scale grassland communities to achieve desired diversity and productivity responses.