Biology subjects

Elofsson, A.

Publications and source records attributed to Elofsson, A..

At least 37 records · Page 2Linked to original sources

bioRxiv2022

Predicting the structure of large protein complexes using AlphaFold and sequential assembly

AlphaFold can predict the structure of single- and multiple-chain proteins with very high accuracy. However, the accuracy decreases with the number of chains, and the available GPU memory limits the size of protein complexes which can be predicted. Here we show that one can predict the structure of large complexes starting from predictions of subcomponents. We assemble 91 out of 175 complexes with 10-30 chains from predicted subcomponents using Monte Carlo tree search, with a median TM-score of 0.51. There are 30 highly accurate complexes (TM-score [≥]0.8, 33% of complete assemblies). We create a scoring function, mpDockQ, that can distinguish if assemblies are complete and predict their accuracy. We find that complexes containing symmetry are accurately assembled, while asymmetrical complexes remain challenging. The method is freely available and accesible as a Colab notebook https://colab.research.google.com/github/patrickbryant1/MoLPC/blob/master/MoLPC.ipynb.

bioinformatics↗

bioRxiv2021

Towards a structurally resolved human protein interaction network

All cellular functions are governed by complex molecular machines that assemble through protein-protein interactions. Their atomic details are critical to the study of their molecular mechanisms but fewer than 5% of hundreds of thousands of human interactions have been structurally characterized. Here, we test the potential and limitations of recent progress in deep-learning methods using AlphaFold2 to predict structures for 65,484 human interactions. We show that higher confidence models are enriched in interactions supported by affinity or structure-based methods and can be orthogonally confirmed by spatial constraints defined by cross-link data. We identify 3,137 high confidence models, of which 1,371 have no homology to a known structure, from which we identify interface residues harbouring disease mutations, suggesting potential mechanisms for pathogenic variants. We find groups of interface phosphorylation sites that show patterns of co-regulation across conditions, suggestive of coordinated tuning of multiple interactions as signalling responses. Finally, we provide examples of how the predicted binary complexes can be used to build larger assemblies. Accurate prediction of protein complexes promises to greatly expand our understanding of the atomic details of human cell biology in health and disease.

bioinformatics↗

bioRxiv2021

A structural biology community assessment of AlphaFold 2 applications

Most proteins fold into 3D structures that determine how they function and orchestrate the biological processes of the cell. Recent developments in computational methods have led to protein structure predictions that have reached the accuracy of experimentally determined models. While this has been independently verified, the implementation of these methods across structural biology applications remains to be tested. Here, we evaluate the use of AlphaFold 2 (AF2) predictions in the study of characteristic structural elements; the impact of missense variants; function and ligand binding site predictions; modelling of interactions; and modelling of experimental structural data. For 11 proteomes, an average of 25% additional residues can be confidently modelled when compared to homology modelling, identifying structural features rarely seen in the PDB. AF2-based predictions of protein disorder and protein complexes surpass state-of-the-art tools and AF2 models can be used across diverse applications equally well compared to experimentally determined structures, when the confidence metrics are critically considered. In summary, we find that these advances are likely to have a transformative impact in structural biology and broader life science research.

biophysics↗

bioRxiv2021

Improved prediction of protein-protein interactions using AlphaFold2 and extended multiple-sequence alignments

Predicting the structure of interacting protein chains is a fundamental step towards understanding protein function. Unfortunately, no computational method can produce accurate structures of protein complexes. AlphaFold2, has shown unprecedented levels of accuracy in modelling single chain protein structures. Here, we apply AlphaFold2 for the prediction of heterodimeric protein complexes. We find that the AlphaFold2 protocol together with optimized multiple sequence alignments, generate models with acceptable quality (DockQ[≥]0.23) for 63% of the dimers. From the predicted interfaces we create a simple function to predict the DockQ score which distinguishes acceptable from incorrect models as well as interacting from non-interacting proteins with state-of-art accuracy. We find that, using the predicted DockQ scores, we can identify 51% of all interacting pairs at 1% FPR. The protocol can be found at: https://gitlab.com/ElofssonLab/FoldDock.

bioinformatics↗

bioRxiv2021

Improved protein docking by predicted interface residues

Scoring docking solutions is a difficult task, and many methods have been developed for this purpose. In docking, only a handful of the hundreds of thousands of models generated by docking algorithms are acceptable, causing difficulties when developing scoring functions. Todays best scoring functions can significantly increase the number of top-ranked models but still fails for most targets. Here, we examine the possibility of utilising predicted residues on a protein-protein interface to score docking models generated during the scan stage of a docking algorithm. Many methods have been developed to infer the portions of a protein surface that interact with another protein, but most have not been benchmarked using docking algorithms. Different interface prediction methods are systematically tested for scoring >300.000 low-resolution rigid-body template free docking decoys. Overall we find that BIPSPI is the best method to identify interface amino acids and score docking solutions. Further, using BIPSPI provides better docking results than state of the art scoring functions, with >12% of first ranked docking models being acceptable. Additional experiments indicated precision as a high-importance metric when estimating interface prediction quality, focusing on docking constraints production. We also discussed several limitations for the adoption of interface predictions as constraints in a docking protocol.

bioinformatics↗

bioRxiv2021

Limits and potential of combined folding and docking using PconsDock.

In the last decade, de novo protein structure prediction accuracy for individual proteins has improved significantly by utilising deep learning (DL) methods for harvesting the co-evolution information from large multiple sequence alignments (MSA). In CASP14, the best groups predicted the structure of most proteins with impressive accuracy. The same approach can, in principle, also be used to extract information about evolutionary-based contacts across protein-protein interfaces. However, most of the earlier studies have not used the latest DL methods for inter-chain contact distance prediction. This paper introduces a fold-and-dock method, PconsDock, based on predicted residue-residue distances with trRosetta. PconsDock can simultaneously predict the tertiary and quaternary structure of a protein pair, even when the structures of the monomers are not known. The straightforward application of this method to a standard dataset for protein-protein docking yielded limited success. However, using alternative methods for MSA generating allowed us to dock accurately significantly more proteins. We also introduced a novel scoring function, PconsDock, that accurately separates 98% of correctly and incorrectly folded and docked proteins. The average performance of the method is comparable to the use of traditional, template-based or ab initio shape-complementarity-only docking methods. However, no a priori structural information for the individual proteins is needed. Moreover, the results of conventional and fold-and-dock approaches are complementary, and thus a combined docking pipeline could increase overall docking success significantly. PconsDocck contributed to the best model for one of the CASP14 oligomeric targets, H1065.

bioinformatics↗

bioRxiv2021

Intra-helical salt bridge contribution to membrane protein insertion

Salt bridges between negatively (D, E) and positively charged (K, R, H) amino acids play an important role in protein stabilization. This has a more prevalent effect in membrane proteins where polar amino acids are exposed to a very hydrophobic environment. In transmembrane (TM) helices the presence of charged residues can hinder the insertion of the helices into the membrane. This can sometimes be avoided by TM region rearrangements after insertion, but it is also possible that the formation of salt bridges could decrease the cost of membrane integration. However, the presence of intra-helical salt bridges in TM domains and their effect on insertion has not been properly studied yet. In this work, we use an analytical pipeline to study the prevalence of charged pairs of amino acid residues in TM -helices, which shows that potentially salt-bridge forming pairs are statistically over-represented. We then selected some candidates to experimentally determine the contribution of these electrostatic interactions to the translocon-assisted membrane insertion process. Using both in vitro and in vivo systems, we confirm the presence of intra-helical salt bridges in TM segments during biogenesis and determined that they contribute between 0.5-0.7 kcal/mol to the apparent free energy of membrane insertion ({Delta}Gapp). Our observations suggest that salt bridge interactions can be stabilized during translocon-mediated insertion and thus could be relevant to consider for the future development of membrane protein prediction software.

biochemistry↗

bioRxiv2021

pyconsFold: A fast and easy tool for modelling and docking using distance predictions.

MotivationContact predictions within a protein has recently become a viable method for accurate prediction of protein structure. Using predicted distance distributions has been shown in many cases to be superior to only using a binary contact annotation. Using predicted inter-protein distances has also been shown to be able to dock some protein dimers. ResultsHere we present pyconsFold. Using CNS as its underlying folding mechanism and predicted contact distance it outperforms regular contact prediction based modelling on our dataset of 210 proteins. It performs marginally worse than the state of the art pyRosetta folding pipeline but is on average about 20 times faster per model. More importantly pyconsFold can also be used as a fold-and-dock protocol by using predicted inter-protein contacts to simultaneously fold and dock two protein chains. Availability and implementationpyconsFold is implemented in Python 3 with a strong focus on using as few dependencies as possible for longevity. It is available both as a pip package in Python 3 and as source code on GitHub and is published under the GPLv3 license. Contactarne@bioinfo.se Supplemental materialInstall instructions, examples and parameters can be found in the supplemental notes. Availability of dataThe data underlying this article together with source code are available on github, at https://github.com/johnlamb/pyconsfold.

bioinformatics↗

bioRxiv2021

The relationship between ageing and changes in the human blood and brain methylomes

Changes in DNA methylation have been found to be strongly correlated with age, enabling the creation of "epigenetic clocks". Previously, studies on the relationship between ageing and DNA methylation have assumed a linear relationship, but no study has shown this always to be the case. We show that the relationships between significant methylation changes and ageing are different in different tissues, and that these changes show variable rates across life. Further, we observe a tendency for saturation as ageing proceeds. We provide a straightforward method of assessing all methylation-age relationships and cluster them according to their relative change rates based on fold change. Less than 13% of the significant markers selected here are used in the most common epigenetic clocks. When the significant markers display largely non-linear relationships, our fold change selection outperforms the most common epigenetic clocks in predicting age. We also show that the saturation observed at older ages explains the earlier observations that the epigenetic clock consistently underestimates the age in older samples. The findings imply that markers selected from linear adjustments or correlations do not represent the most meaningful biological changes.

bioinformatics↗

bioRxiv2020

The evolutionary history of topological variations in the CPA/AT superfamily

CPA/AT transporters consist of two structurally and evolutionarily related inverted repeat units, each of them with one core and one scaffold subdomain. During evolution, these families have undergone substantial changes in structure, topology and function. Central to the function of the transporters is the existence of two non-canonical helices that are involved in the transport process. In different families, two different types of these helices have been identified, reentrant and broken. Here, we use an integrated topology annotation method to identify novel topologies in the families. It combines topology prediction, similarity to families with known structure, and the difference in positively charged residues present in inside and outside loops in alternative topological models. We identified families with diverse topologies containing broken or reentrant helix. We classified all families based on 3 distinct evolutionary groups that each share a structurally similar C-terminal repeat unit newly termed as "Fold-types". Using the evolutionary relationship between families we propose topological transitions including, a transition between broken and reentrant helices, complete change of orientation, changes in the number of scaffold helices and even in some rare cases, losses of core helices. The evolutionary history of the repeat units shows gene duplication and repeat shuffling events to result in these extensive topology variations. The novel structure-based classification, together with supporting structural models and other information, is presented in a searchable database, CPAfold (cpafold.bioinfo.se). Our comprehensive study of topology variations within the CPA superfamily provides better insight about their structure and evolution.

bioinformatics↗

bioRxiv2020

Identification of a novel fold type in CPA/AT transporters by ab-initio structure prediction

Members of the CPA/AT transporter superfamily show significant structural variability. All previously known members consist of an inverted duplicated repeat unit that folds into two separate domains, the core and the scaffold domain. Crucial for its transporting function, the central helix in the core domain is a noncanonical transmembrane helix, which can either be in the form of a broken helix or a reentrant helix. Here, we expand the structural knowledge of the CPA/AT family by using contact-prediction-based protein modelling. We show that the N-terminal domains of the Pfam families; PSE (Cons_hypoth698 PF03601), Lysine exporter (PF03956) and LrgB (PF04172) families have a previously unseen reentrant-helix-reentrant fold. The close homology between PSE and the Sodium-citrate symporter (2HCT) suggests that the new fold originates from the truncation of an ancestral reentrant protein, caused by the loss of the C-terminal reentrant helix. To compensate for the lost reentrant helix one external loop moves into the membrane to form the second reentrant helix, highlighting the adaptability of the CPA/AT transporters. This study also demonstrates that the most recent deep-learning-based modelling methods have become a useful tool to gain biologically relevant structural, evolutionary and functional insights about protein families.

bioinformatics↗

bioRxiv2019

The Use of GC-, Codon-, and Amino Acid-frequencies to Understand the Evolutionary Forces at a Genomic Scale.

1It is well known that the GC content varies enormously between organisms; this is believed to be caused by a combination of mutational preferences and selective pressure. Within coding regions, the variation of GC is more substantial in position three and smaller in position one and two. Less well known is that this variation also has an enormous impact on the frequency of amino acids as their codons vary in GC content. For instance, the fraction of alanines in different proteomes varies from 1.1% to 16.5%. In general, the frequency of different amino acids correlates strongly with the number of codons, the GC content of these codons and the genomic GC contents. However, there are clear and systematic deviations from the expected frequencies. Some amino acids are more frequent than expected by chance, while others are less frequent. A plausible model to explain this is that there exist two different selective forces acting on the genes; First, there exists a force acting to maintain the overall GC level and secondly there exists a selective force acting on the amino acid level. Here, we use the divergence in amino acid frequency from what is expected by the GC content to analyze the selective pressure acting on codon frequencies in the three kingdoms of life. We find four major selective forces; First, the frequency of serine is lower than expected in all genomes, but most in prokaryotes. Secondly, there exist a selective pressure acting to balance positively and negatively charged amino acids, which results in a reduction of arginine and negatively charged amino acids. This results in a reduction of arginine and all the negatively charged amino acids. Thirdly, the frequency of the hydrophobic residues encoded by a T in the second codon position does not change with GC. Their frequency is lower in eukaryotes than in prokaryotes. Finally, some amino acids with unique properties, such as proline glycine and proline, are limited in their frequency variation.

evolutionary biology↗

bioRxiv2019

Accurate contact-based modelling of repeat proteins predicts the structure of Curlin and SPW repeats.

Repeat proteins are an abundant class in eukaryotic proteomes. They are involved in many eukaryotic specific functions, including signalling. For many of these families, the structure is not known. Recently, it has been shown that the structure of many protein families can be predicted by using contact predictions from direct coupling analysis and deep learning. However, their unique sequence features present in repeat proteins is a challenge for contact predictions DCA-methods. Here, we show that using the deep learning-based PconsC4 is more effective for predicting both intra and interunit contacts among a comprehensive set of repeat proteins. In a benchmark dataset of 819 repeat proteins about one third can be correctly modelled and among 51 PFAM families lacking a protein structure, we produce models of five families with estimated high accuracy.\n\nAuthor SummaryRepeat proteins are widespread among organisms and particularly abundant in eukaryotic proteomes. Their primary sequence present repetition in the amino acid sequences that origin structures with repeated folds/domains. Although the repeated units are easy to be recognized in primary sequence, often structure information are missing. Here we used contact prediction for predicting the structure of repeats protein directly from their primary sequences. We benchmark our method on a dataset comprehensive of all the known repeated structures. We evaluate the contact predictions and the obtained models set for different classes of proteins and different lengths of the target, and we benchmark the quality assessment of the models on repeats proteins. Finally, we applied the methods on the repeat PFAM families missing of resolved structures, five of them modelled with high accuracy.

bioinformatics↗

bioRxiv2019

Detecting Novel Sequence Signals in Targeting Peptides Using Deep Learning

In bioinformatics, machine learning methods have been used to predict features embedded in the sequences. In contrast to what is generally assumed, machine learning approaches can also provide new insights into the underlying biology. Here, we demonstrate this by presenting TargetP 2.0, a novel state of art method to identify N-terminal sorting signals, which direct proteins to the secretory pathway, mitochondria and chloroplasts or other plastids.\n\nBy examining the strongest signals from the attention layer in the network, we find that the second residue in the protein, i.e. the one following the initial methionine, has a strong influence on the classification. When subsequently examining all targeting peptides, we observe that two-thirds of chloroplast and thylakoid transit peptides have an alanine in position two, but only 20% of other plant proteins. Further highlighting the importance of the second residue, we also note that in fungi and single-celled eukaryotes, less than 30% of the targeting peptides have an amino acid that allows the removal of the N-terminal methionine compared with 60% for the proteins without targeting peptide.\n\nTargetP 2.0 is available at http://www.cbs.dtu.dk/services/TargetP-2.0/index.php

bioinformatics↗

bioRxiv2018

PconsC4: fast, free, easy, and accurate contact predictions.

MotivationResidue contact prediction was revolutionized recently by the introduction of direct coupling analysis (DCA). Further improvements, in particular for small families, have been obtained by the combination of DCA and deep learning methods. However, existing deep learning contact prediction methods often rely on a number of external programs and are therefore computationally expensive.\n\nResultsHere, we introduce a novel contact predictor, PconsC4, which performs on par with state of the art methods. PconsC4 is heavily optimized, does not use any external programs and therefore is significantly faster and easier to use than other methods.\n\nAvailabilityPconsC4 is freely available under the GPL license from https://github.com/ElofssonLab/PconsC4. Installation is easy using the pip command and works on any system with Python 3.5 or later and a modern GCC compiler.\n\nContactarne@bioinfo.se

bioinformatics↗

bioRxiv2018

Improved protein model quality assessments by changing the target function

Protein modeling quality is an important part of protein structure prediction. We have for more than a decade developed a set of methods for this problem. We have used various types of description of the protein and different machine learning methodologies. However, common to all these methods has been the target function used for training. The target function in ProQ describes the local quality of a residue in a protein model. In all versions of ProQ the target function has been the S-score. However, other quality estimation functions also exist, which can be divided into superposition- and contact-based methods. The superposition-based methods, such as S-score, are based on a rigid body superposition of a protein model and the native structure, while the contact-based methods compare the local environment of each residue. Here, we examine the effects of retraining our latest predictor, ProQ3D, using identical inputs but different target functions. We find that the c ntact-based methods are easier to predict and that predictors trained on these measures provide some advantages when it comes to identifying the best model. One possible reason for this is that contact based methods are better at estimating the quality of multi-domain targets. However, training on the S-score gives the best correlation with the GDT_TS score, which is commonly used in CASP to score the global model quality. To take the advantage of both of these features we provide an updated version of ProQ3D that predicts local and global model quality estimates based on different quality estimates.

bioinformatics↗

bioRxiv2018

Why do eukaryotic proteins contain more intrinsically disordered regions?

Intrinsic disorder is much more abundant in eukaryotic than in prokaryotic proteins. However, the reason behind this is unclear. It has been proposed that the disordered regions are functionally important for regulation in eukaryotes, but it has also been proposed that the difference is a result of lower selective pressure in eukaryotes. Almost all studies intrinsic disorder is predicted from the amino acid sequence of a protein. Therefore, there should exist an underlying difference in the amino acid distributions between eukaryotic and prokaryotic proteins causing the predicted difference in intrinsic disorder. To obtain a better understanding of why eukaryotic proteins contain more intrinsically disordered regions we compare proteins from complete eukaryotic and prokaryotic proteomes.\n\nHere, we show that the difference in intrinsic disorder origin from differences in the linker regions. Eukaryotic proteins have more extended linker regions and, in particular, the eukaryotic linker regions are more disordered. The average eukaryotic protein is about 500 residues long; it contains 250 residues in linker regions, of which 80 are disordered. In comparison, prokaryotic proteins are about 350 residues long and only have 100-110 residues in linker regions, and less than 10 of these are intrinsically disordered.\n\nFurther, we show that there is no systematic increase in the frequency of disorder-promoting residues in eukaryotic linker regions. Instead, the difference in frequency of only three amino acids seems to lie behind the difference. The most significant difference is that eukaryotic linkers contain about 9% serine, while prokaryotic linkers have roughly 6.5%. Eukaryotic linkers also contain about 2% more proline and 2-3% fewer isoleucine residues. The reason why primarily these amino acids vary in frequency is not apparent, but it cannot be excluded that the difference is serine is related to the increased need for regulation through phosphorylation and that the proline difference is related to increase of eukaryotic specific repeats.

bioinformatics↗

bioRxiv2017

The number of orphans in yeast and fly is drastically reduced by using combining searches in both proteomes and genomes

The identification of de novo created genes is important as it provides a glimpse on the evolutionary processes of gene creation. Potential de novo created genes are identified by selecting genes that have no homologs outside a particular species, but for an accurate detection this identification needs to be correct.\n\nGenes without any homologs are often referred to as orphans; in addition to de novo created ones, fast evolving genes or genes lost in all related genomes might also be classified as orphans. The identification of orphans is dependent on: (i) a method to detect homologs and (ii) a database including genes from related genomes.\n\nHere, we set out to investigate how the detection of orphans is influenced by these two factors. Using Saccharomyces cerevisiae we identify that best strategy is to use a combination of searching annotated proteins and a six-frame translation of all ORFs from closely related genomes. Using this strategy we obtain a set of 54 orphans in Drosophila melanogaster and 38 in Drosophila pseudoobscura, significantly less than what is reported in some earlier studies.

bioinformatics↗