The process of heme binding to a protein is prevalent in almost all forms of life to control many important biological properties, such as O2-binding, electron transfer, gas sensing or to build catalytic power. In these cases, heme typically binds tightly (irreversibly) to a protein in a discrete heme binding pocket, with one or two heme ligands provided most commonly to the heme iron by His, Cys or Tyr residues. Heme binding can also be used as a regulatory mechanism, for example in transcriptional regulation or ion channel control. When used as a regulator, heme binds more weakly, with different heme ligations and without the need for a discrete heme pocket. This makes characterisation of heme regulatory proteins difficult, and new approaches are needed to predict and understand the heme-protein interactions. We apply a modified version of the ProFunc bioinformatics tool to identify heme binding sites in a test set of heme-dependent regulatory proteins taken from the Protein Data Bank (PDB) and AlphaFold models. The potential heme binding sites identified can be easily visualised in PyMol and, if necessary, optimised with RosettaDOCK. We demonstrate that the methodology can be used to identify heme binding sites in proteins, including in cases where there is no crystal structure available, but the methodology is more accurate when the quality of the structural information is high. The ProFunc tool, with the modification used in this work, is publicly available at https://www.ebi.ac.uk/thornton-srv/databases/profunc and can be readily adopted for the examination of new heme binding targets.

Rare genetic diseases affect millions, and identifying causal DNA variants is essential for patient care. Therefore, it is imperative to estimate the effect of each independent variant and improve their pathogenicity classification. Our study of 140 214 unrelated UK Biobank (UKB) participants found that each of them carries a median of 7 variants previously reported as pathogenic or likely pathogenic. We focused on 967 diagnostic-grade gene (DGG) variants for rare bleeding, thrombotic, and platelet disorders (BTPDs) observed in 12 367 UKB participants. By association analysis, for a subset of these variants, we estimated effect sizes for platelet count and volume, and odds ratios for bleeding and thrombosis. Variants causal of some autosomal recessive platelet disorders revealed phenotypic consequences in carriers. Loss-of-function variants in MPL, which cause chronic amegakaryocytic thrombocytopenia if biallelic, were unexpectedly associated with increased platelet counts in carriers. We also demonstrated that common variants identified by genome-wide association studies (GWAS) for platelet count or thrombosis risk may influence the penetrance of rare variants in BTPD DGGs on their associated hemostasis disorders. Network-propagation analysis applied to an interactome of 18 410 nodes and 571 917 edges showed that GWAS variants with large effect sizes are enriched in DGGs and their first-order interactors. Finally, we illustrate the modifying effect of polygenic scores for platelet count and thrombosis risk on disease severity in participants carrying rare variants in TUBB1 or PROC and PROS1, respectively. Our findings demonstrate the power of association analyses using large population datasets in improving pathogenicity classifications of rare variants.

Constrained Coding Regions (CCRs) in the human genome have been derived from DNA sequencing data of large cohorts of healthy control populations, available in the Genome Aggregation Database (gnomAD) [1]. They identify regions depleted of protein-changing variants and thus identify segments of the genome that have been constrained during human evolution. By mapping these DNA-defined regions from genomic coordinates onto the corresponding protein positions and combining this information with protein annota-tions, we have explored the distribution of CCRs and compared their co-occurrence with different protein functional features, previously annotated at the amino acid level in public databases. As expected, our results reveal that functional amino acids involved in interactions with DNA/RNA, protein-protein contacts and catalytic sites are the protein features most likely to be highly constrained for variation in the control pop-ulation. More surprisingly, we also found that linear motifs, linear interacting peptides (LIPs), disorder-order transitions upon binding with other protein partners and liquid-liquid phase separating (LLPS) regions are also strongly associated with high constraint for variability. We also compared intra-species constraints in the human CCRs with inter-species conservation and functional residues to explore how such CCRs may contribute to the analysis of protein variants. As has been previously observed, CCRs are only weakly cor-related with conservation, suggesting that intraspecies constraints complement interspecies conservation and can provide more information to interpret variant effects.

Most proteins fold into 3D structures that determine how they function and orchestrate the biological processes of the cell. Recent developments in computational methods for protein structure predictions have reached the accuracy of experimentally determined models. Although this has been independently verified, the implementation of these methods across structural-biology applications remains to be tested. Here, we evaluate the use of AlphaFold2 (AF2) predictions in the study of characteristic structural elements; the impact of missense variants; function and ligand binding site predictions; modeling of interactions; and modeling of experimental structural data. For 11 proteomes, an average of 25% additional residues can be confidently modeled when compared with homology modeling, identifying structural features rarely seen in the Protein Data Bank. AF2-based predictions of protein disorder and complexes surpass dedicated tools, and AF2 models can be used across diverse applications equally well compared with 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.

PDBsum1 is a standalone set of programs to perform the same structural analyses as provided by the PDBsum web server (https://www.ebi.ac.uk/pdbsum). The server has pages for every entry in the Protein Data Bank (PDB) and can also process user-uploaded PDB files, returning a password-protected set of pages that are retained for around 3 months. The standalone version described here allows for in-house processing and indefinite retention of the results. All data files and images are pre-generated, rather than on-the-fly as in the web version, so can be easily accessed. The program runs on Linux, Windows, and mac operating systems and is freely available for academic use at https://www.ebi.ac.uk/thornton-srv/software/PDBsum1.

DNA-Damage Response (DDR) proteins are crucial for maintaining the integrity of the genome by identifying and repairing errors in DNA. Variants affecting their function can have severe consequences since failure to repair damaged DNA can result in cells turning cancerous. Here, we compare germline and somatic variants in DDR genes, specifically looking at their locations in the corresponding three-dimensional (3D) structures, Pfam domains, and protein–protein interaction interfaces. We show that somatic variants in metastatic cases are more likely to be found in Pfam domains and protein interaction interfaces than are pathogenic germline variants or variants of unknown significance (VUS). We also show that there are hotspots in the structures of ATM and BRCA2 proteins where pathogenic germline, and recurrent somatic variants from primary and metastatic tumours, cluster together in 3D. Moreover, in the ATM, BRCA1 and BRCA2 genes from prostate cancer patients, the distributions of germline benign, pathogenic, VUS, and recurrent somatic variants differ across Pfam domains. Together, these results provide a better characterisation of the most recurrent affected regions in DDRs and could help in the understanding of individual susceptibility to tumour development.

Protein structures predicted using artificial intelligence will aid medical research, but the greatest benefit will come if clinical data can be similarly used to better understand human disease.

The PDBsum web server provides structural analyses of the entries in the Protein Data Bank (PDB). Two recent additions are described here. The first is the detailed analysis of the SARS-CoV-2 virus protein structures in the PDB. These include the variants of concern, which are shown both on the sequences and 3D structures of the proteins. The second addition is the inclusion of the available AlphaFold models for human proteins. The pages allow a search of the protein against existing structures in the PDB via the Sequence Annotated by Structure (SAS) server, so one can easily compare the predicted model against experimentally determined structures. The server is freely accessible to all at http://www.ebi.ac.uk/pdbsum.

The prediction of peptide binders to Major Histocompatibility Complex (MHC) class II receptors is of great interest to study autoimmune diseases and for vaccine development. Most approaches predict the affinities using sequence-based models trained on experimental data and multiple alignments from known peptide substrates. However, detecting activity differences caused by single-point mutations is a challenging task. In this work, we used interactions calculated from simulations to build scoring matrices for quickly estimating binding differences by single-point mutations. We modelled a set of 837 peptides bound to an MHC class II allele, and optimized the sampling of the conformations using the Rosetta backrub method by comparing the results to molecular dynamics simulations. From the dynamic trajectories of each complex, we averaged and compared structural observables for each amino acid at each position of the 9°mer peptide core region. With this information, we generated the scoring-matrices to predict the sign of the binding differences. We then compared the performance of the best scoring-matrix to different computational methodologies that range in computational costs. Overall, the prediction of the activity differences caused by single mutated peptides was lower than 60% for all the methods. However, the developed scoring-matrix in combination with existing methods reports an increase in the performance, up to 86% with a scoring method that uses molecular dynamics.

Background Proteases are key drivers in many biological processes, in part due to their specificity towards their substrates. However, depending on the family and molecular function, they can also display substrate promiscuity which can also be essential. Databases compiling specificity matrices derived from experimental assays have provided valuable insights into protease substrate recognition. Despite this, there are still gaps in our knowledge of the structural determinants. Here, we compile a set of protease crystal structures with bound peptide-like ligands to create a protocol for modelling substrates bound to protease structures, and for studying observables associated to the binding recognition. Results As an application, we modelled a subset of protease–peptide complexes for which experimental cleavage data are available to compare with informational entropies obtained from protease–specificity matrices. The modelled complexes were subjected to conformational sampling using the Backrub method in Rosetta, and multiple observables from the simulations were calculated and compared per peptide position. We found that some of the calculated structural observables, such as the relative accessible surface area and the interaction energy, can help characterize a protease’s substrate recognition, giving insights for the potential prediction of novel substrates by combining additional approaches. Conclusion Overall, our approach provides a repository of protease structures with annotated data, and an open source computational protocol to reproduce the modelling and dynamic analysis of the protease–peptide complexes.

The geometry of atomic arrangement underpins the structural understanding of molecules in many fields. However, no general framework of mathematical/computational theory for the geometry of atomic arrangement exists. Here we present "Molecular Geometry (MG)" as a theoretical framework accompanied by "MG Operating System (MGOS)" which consists of callable functions implementing the MG theory. MG allows researchers to model complicated molecular structure problems in terms of elementary yet standard notions of volume, area, etc. and MGOS frees them from the hard and tedious task of developing/implementing geometric algorithms so that they can focus more on their primary research issues. MG facilitates simpler modeling of molecular structure problems; MGOS functions can be conveniently embedded in application programs for the efficient and accurate solution of geometric queries involving atomic arrangements. The use of MGOS in problems involving spherical entities is akin to the use of math libraries in general purpose programming languages in science and engineering.

MOTIVATION:Understanding the protein structural context and patterning on proteins of genomic variants can help to separate benign from pathogenic variants and reveal molecular consequences. However, mapping genomic coordinates to protein structures is non-trivial, complicated by alternative splicing and transcript evidence.RESULTS:Here we present VarMap, a web tool for mapping a list of chromosome coordinates to canonical UniProt sequences and associated protein 3D structures, including validation checks, and annotating them with structural information.AVAILABILITY AND IMPLEMENTATION:https://www.ebi.ac.uk/thornton-srv/databases/VarMap.SUPPLEMENTARY INFORMATION:Supplementary data are available at Bioinformatics online.

VarSite is a web server mapping known disease-associated variants from UniProt and ClinVar, together with natural variants from gnomAD, onto protein 3D structures in the Protein Data Bank. The analyses are primarily image-based and provide both an overview for each human protein, as well as a report for any specific variant of interest. The information can be useful in assessing whether a given variant might be pathogenic or benign. The structural annotations for each position in the protein include protein secondary structure, interactions with ligand, metal, DNA/RNA, or other protein, and various measures of a given variant's possible impact on the protein's function. The 3D locations of the disease-associated variants can be viewed interactively via the 3dmol.js JavaScript viewer, as well as in RasMol and PyMOL. Users can search for specific variants, or sets of variants, by providing the DNA coordinates of the base change(s) of interest. Additionally, various agglomerative analyses are given, such as the mapping of disease and natural variants onto specific Pfam or CATH domains. The server is freely accessible to all at: https://www.ebi.ac.uk/thornton-srv/databases/VarSite.

The ProFunc web server is a tool for helping identify the function of a given protein whose 3D coordinates have been experimentally determined or homology modeled. It uses a cocktail of both sequence- and structure-based methods to identify matches to other proteins that may, in turn, suggest the query protein's most likely function. The server was originally developed to aid the worldwide structural genomics effort at the start of the millennium. It accepts a file containing the protein's 3D coordinates in PDB format, and, when processing is complete, sends an email containing a link to the password-protected result pages. The results include an at-a-glance summary, as well as separate pages containing more detailed analyses. The server can be found at: http://www.ebi.ac.uk/thornton-srv/databases/profunc .

BackgroundSyntaxin-binding protein 1, encoded by STXBP1, is highly expressed in the brain and involved in fusing synaptic vesicles with the plasma membrane. Studies have shown that pathogenic loss-of-function variants in this gene result in various types of epilepsies, mostly beginning early in life. We were interested to model pathogenic missense variants on the protein structure to investigate the mechanism of pathogenicity and genotype-phenotype correlations. MethodsWe report 11 patients with pathogenic de novo mutations in STXBP1 identified in the first 4293 trios of the Deciphering Developmental Disorder (DDD) study, including six missense variants. We analyzed the structural locations of the pathogenic missense variants from this study and the literature, as well as population missense variants extracted from Exome Aggregation Consortium (ExAC). ResultsPathogenic variants are significantly more likely to occur at highly conserved locations than population variants, and be buried inside the protein domain. Pathogenic mutations are also more likely to destabilize the domain structure compared with population variants, increasing the proportion of (partially) unfolded domains that are prone to aggregation or degradation. We were unable to detect any genotype-phenotype correlation, but unlike previously reported cases, most of the DDD patients with STXBP1 pathogenic variants did not present with very early-onset or severe epilepsy and encephalopathy, though all have developmental delay with intellectual disability and most display behavioral problems and suffered seizures in later childhood. ConclusionVariants across STXBP1 that cause loss of function can result in severe intellectual disability with or without seizures, consistent with a haploinsufficiency mechanism. Pathogenic missense mutations act through destabilization of the protein domain, making it prone to aggregation or degradation. The presence or absence of early seizures may reflect ascertainment bias in the literature as well as the broad recruitment strategy of the DDD study.

Haploinsufficiency in DYRK1A is associated with a recognizable developmental syndrome, though the mechanism of action of pathogenic missense mutations is currently unclear. Here we present 19 de novo mutations in this gene, including five missense mutations, identified by the Deciphering Developmental Disorder study. Protein structural analysis reveals that the missense mutations are either close to the ATP or peptide binding-sites within the kinase domain, or are important for protein stability, suggesting they lead to a loss of the protein's function mechanism. Furthermore, there is some correlation between the magnitude of the change and the severity of the resultant phenotype. A comparison of the distribution of the pathogenic mutations along the length of DYRK1A with that of natural variants, as found in the ExAC database, confirms that mutations in the N-terminal end of the kinase domain are more disruptive of protein function. In particular, pathogenic mutations occur in significantly closer proximity to the ATP and the substrate peptide than the natural variants. Overall, we suggest that de novo dominant mutations in DYRK1A account for nearly 0.5% of severe developmental disorders due to substantially reduced kinase function.

Decades of intensive experimental studies of the recognition of DNA sequences by proteins have provided us with a view of a diverse and complicated world in which few to no features are shared between individual DNA-binding protein families. The originally conceived direct readout of DNA residue sequences by amino acid side chains offers very limited capacity for sequence recognition, while the effects of the dynamic properties of the interacting partners remain difficult to quantify and almost impossible to generalise. In this work we investigated the energetic characteristics of all DNA residue—amino acid side chain combinations in the conformations found at the interaction interface in a very large set of protein—DNA complexes by the means of empirical potential-based calculations. General specificity-defining criteria were derived and utilised to look beyond the binding motifs considered in previous studies. Linking energetic favourability to the observed geometrical preferences, our approach reveals several additional amino acid motifs which can distinguish between individual DNA bases. Our results remained valid in environments with various dielectric properties.

We present a generic, multidisciplinary approach for improving our understanding of novel missense variants in recently discovered disease genes exhibiting genetic heterogeneity, by combining clinical and population genetics with protein structural analysis. Using six new de novo missense diagnoses in TBL1XR1 from the Deciphering Developmental Disorders study, together with population variation data, we show that the β-propeller structure of the ubiquitous WD40 domain provides a convincing way to discriminate between pathogenic and benign variation. Children with likely pathogenic mutations in this gene have severely delayed language development, often accompanied by intellectual disability, autism, dysmorphology and gastrointestinal problems. Amino acids affected by likely pathogenic missense mutations are either crucial for the stability of the fold, forming part of a highly conserved symmetrically repeating hydrogen-bonded tetrad, or located at the top face of the β-propeller, where ‘hotspot’ residues affect the binding of β-catenin to the TBLR1 protein. In contrast, those altered by population variation are significantly less likely to be spatially clustered towards the top face or to be at buried or highly conserved residues. This result is useful not only for interpreting benign and pathogenic missense variants in this gene, but also in other WD40 domains, many of which are associated with disease.