mCSM-lig: quantifying the effects of mutations on protein-small molecule affinity in genetic disease and emergence of drug resistance

The ability to predict how a mutation affects ligand binding is an essential step in understanding, anticipating and improving the design of new treatments for drug resistance, and in understanding genetic diseases. Here we present mCSM-lig, a structure-guided computational approach for quantifying the effects of single-point missense mutations on affinities of small molecules for proteins. mCSM-lig uses graph-based signatures to represent the wild-type environment of mutations, and small-molecule chemical features and changes in protein stability as evidence to train a predictive model using a representative set of protein-ligand complexes from the Platinum database. We show our method provides a very good correlation with experimental data (up to ρ = 0.67) and is effective in predicting a range of chemotherapeutic, antiviral and antibiotic resistance mutations, providing useful insights for genotypic screening and to guide drug development. mCSM-lig also provides insights into understanding Mendelian disease mutations and as a tool for guiding protein design. mCSM-lig is freely available as a web server at http://structure.bioc.cam.ac.uk/mcsm_lig.

mCSM-AB: a web server for predicting antibody-antigen affinity changes upon mutation with graph-based signatures

Computational methods have traditionally struggled to predict the effect of mutations in antibody–antigen complexes on binding affinity. This has limited their usefulness during antibody engineering and development, and their ability to predict biologically relevant escape mutations. Here we present mCSM-AB, a user-friendly web server for accurately predicting antibody–antigen affinity changes upon mutation which relies on graph-based signatures. We show that mCSM-AB performs better than comparable methods that have been previously used for antibody engineering. mCSM-AB web server is available at http://structure.bioc.cam.ac.uk/mcsm_ab.

CSM-lig: a web server for assessing and comparing protein-small molecule affinities

Determining the affinity of a ligand for a given protein is a crucial component of drug development and understanding their biological effects. Predicting binding affinities is a challenging and difficult task, and despite being regarded as poorly predictive, scoring functions play an important role in the analysis of molecular docking results. Here, we present CSM-Lig (http://structure.bioc.cam.ac.uk/csm_lig), a web server tailored to predict the binding affinity of a protein-small molecule complex, encompassing both protein and small-molecule complementarity in terms of shape and chemistry via graph-based structural signatures. CSM-Lig was trained and evaluated on different releases of the PDBbind databases, achieving a correlation of up to 0.86 on 10-fold cross validation and 0.80 in blind tests, performing as well as or better than other widely used methods. The web server allows users to rapidly and automatically predict binding affinities of collections of structures and assess the interactions made. We believe CSM-lig would be an invaluable tool for helping assess docking poses, the effects of multiple mutations, including insertions, deletions and alternative splicing events, in protein-small molecule affinity, unraveling important aspects that drive protein–compound recognition.

Mycobacterium tuberculosis whole genome sequencing and protein structure modelling provides insights into anti-tuberculosis drug resistance

BACKGROUND

Combating the spread of drug resistant tuberculosis is a global health priority. Whole genome association studies are being applied to identify genetic determinants of resistance to anti-tuberculosis drugs. Protein structure and interaction modelling are used to understand the functional effects of putative mutations and provide insight into the molecular mechanisms leading to resistance.

METHODS

To investigate the potential utility of these approaches, we analysed the genomes of 144 Mycobacterium tuberculosis clinical isolates from The Special Programme for Research and Training in Tropical Diseases (TDR) collection sourced from 20 countries in four continents. A genome-wide approach was applied to 127 isolates to identify polymorphisms associated with minimum inhibitory concentrations for first-line anti-tuberculosis drugs. In addition, the effect of identified candidate mutations on protein stability and interactions was assessed quantitatively with well-established computational methods.

RESULTS

The analysis revealed that mutations in the genes rpoB (rifampicin), katG (isoniazid), inhA-promoter (isoniazid), rpsL (streptomycin) and embB (ethambutol) were responsible for the majority of resistance observed. A subset of the mutations identified in rpoB and katG were predicted to affect protein stability. Further, a strong direct correlation was observed between the minimum inhibitory concentration values and the distance of the mutated residues in the three-dimensional structures of rpoB and katG to their respective drugs binding sites.

CONCLUSIONS

Using the TDR resource, we demonstrate the usefulness of whole genome association and convergent evolution approaches to detect known and potentially novel mutations associated with drug resistance. Further, protein structural modelling could provide a means of predicting the impact of polymorphisms on drug efficacy in the absence of phenotypic data. These approaches could ultimately lead to novel resistance mutations to improve the design of tuberculosis control measures, such as diagnostics, and inform patient management.

Conjugation of 10 kDa Linear PEG onto Trastuzumab Fab' Is Sufficient to Significantly Enhance Lymphatic Exposure while Preserving in Vitro Biological Activity

The lymphatic system is a major conduit by which many diseases spread and proliferate. There is therefore increasing interest in promoting better lymphatic drug targeting. Further, antibody fragments such as Fabs have several advantages over full length monoclonal antibodies but are subject to rapid plasma clearance, which can limit the lymphatic exposure and activity of Fabs against lymph-resident diseases. This study therefore explored ideal PEGylation strategies to maximize biological activity and lymphatic exposure using trastuzumab Fab' as a model. Specifically, the Fab' was conjugated with single linear 10 or 40 kDa PEG chains at the hinge region. PEGylation led to a 3-4-fold reduction in binding affinity to HER2, but antiproliferative activity against HER2-expressing BT474 cells was preserved. Lymphatic pharmacokinetics were then examined in thoracic lymph duct cannulated rats after intravenous and subcutaneous dosing at 2 mg/kg, and the data were evaluated via population pharmacokinetic modeling. The Fab' displayed limited lymphatic exposure, but conjugation of 10 kDa PEG improved exposure by approximately 11- and 5-fold after intravenous (15% dose collected in thoracic lymph over 30 h) and subcutaneous (9%) administration, respectively. Increasing the molecular weight of the PEG to 40 kDa, however, had no significant impact on lymphatic exposure after intravenous (14%) administration and only doubled lymphatic exposure after subcutaneous administration (18%) when compared to 10 kDa PEG-Fab'. The data therefore suggests that minimal PEGylation has the potential to enhance the exposure and activity of Fab's against lymph-resident diseases, while no significant benefit is achieved with very large PEGs.

In silico functional dissection of saturation mutagenesis: Interpreting the relationship between phenotypes and changes in protein stability, interactions and activity

Despite interest in associating polymorphisms with clinical or experimental phenotypes, functional interpretation of mutation data has lagged behind generation of data from modern high-throughput techniques and the accurate prediction of the molecular impact of a mutation remains a non-trivial task. We present here an integrated knowledge-driven computational workflow designed to evaluate the effects of experimental and disease missense mutations on protein structure and interactions. We exemplify its application with analyses of saturation mutagenesis of DBR1 and Gal4 and show that the experimental phenotypes for over 80% of the mutations correlate well with predicted effects of mutations on protein stability and RNA binding affinity. We also show that analysis of mutations in VHL using our workflow provides valuable insights into the effects of mutations, and their links to the risk of developing renal carcinoma. Taken together the analyses of the three examples demonstrate that structural bioinformatics tools, when applied in a systematic, integrated way, can rapidly analyse a given system to provide a powerful approach for predicting structural and functional effects of thousands of mutations in order to reveal molecular mechanisms leading to a phenotype. Missense or non-synonymous mutations are nucleotide substitutions that alter the amino acid sequence of a protein. Their effects can range from modifying transcription, translation, processing and splicing, localization, changing stability of the protein, altering its dynamics or interactions with other proteins, nucleic acids and ligands, including small molecules and metal ions. The advent of high-throughput techniques including sequencing and saturation mutagenesis has provided large amounts of phenotypic data linked to mutations. However, one of the hurdles has been understanding and quantifying the effects of a particular mutation, and how they translate into a given phenotype. One approach to overcome this is to use robust, accurate and scalable computational methods to understand and correlate structural effects of mutations with disease.

Exploring the chemical space of the lysine-binding pocket of the first kringle domain of hepatocyte growth factor/scatter factor (HGF/SF) yields a new class of inhibitors of HGF/SF-MET binding

The growth/motility factor hepatocyte growth factor/scatter factor (HGF/SF) and its receptor, the tyrosine kinase MET, constitute a signalling system essential for embryogenesis and for tissue/organ regeneration in post-natal life. HGF/SF-MET signalling, however, also plays a key role in the onset of metastasis of a large number of human tumours. Both HGF/SF and MET are high molecular weight proteins that bury an extensive interface upon complex formation and thus constitute a challenging target for the development of low molecular weight inhibitors. Here we have used surface plasmon resonance (SPR), nuclear magnetic resonance (NMR) and X-ray crystallography to screen a diverse fragment library of 1338 members as well as a range of piperazine-like compounds. Several small molecules were found to bind in the lysine-binding pocket of the kringle 1 domain of HGF/SF and its truncated splice variant NK1. We have defined the binding mode of these compounds, explored their biological activity and we show that selected fragments inhibit MET downstream signalling. Thus we demonstrate that targeting the lysine-binding pocket of NK1 is an effective strategy to generate MET receptor antagonists and we offer proof of concept that the HGF/SF-MET interface may be successfully targeted with small molecules. These studies have broad implications for the development of HGF/SF-MET therapeutics and cancer treatment.

Achieving high signal-to-noise in cell regulatory systems: Spatial organization of multiprotein transmembrane assemblies of FGFR and MET receptors

How is information communicated both within and between cells of living systems with high signal to noise? We discuss transmembrane signaling models involving two receptor tyrosine kinases: the fibroblast growth factor receptor (FGFR) and the MET receptor. We suggest that simple dimerization models might occur opportunistically giving rise to noise but cooperative clustering of the receptor tyrosine kinases observed in these systems is likely to be important for signal transduction. We propose that this may be a more general prerequisite for high signal to noise in transmembrane receptor signaling.

pkCSM: Predicting Small-Molecule Pharmacokinetic and Toxicity Properties Using Graph-Based Signatures

Drug development has a high attrition rate, with poor pharmacokinetic and safety properties a significant hurdle. Computational approaches may help minimize these risks. We have developed a novel approach (pkCSM) which uses graph-based signatures to develop predictive models of central ADMET properties for drug development. pkCSM performs as well or better than current methods. A freely accessible web server (http://structure.bioc.cam.ac.uk/pkcsm), which retains no information submitted to it, provides an integrated platform to rapidly evaluate pharmacokinetic and toxicity properties.

PEGylated interferon displays differences in plasma clearance and bioavailability between male and female mice and between female immunocompetent C57Bl/6J and athymic nude mice

Gender and immune status can considerably impact on the pharmacokinetics (PK) of macromolecular and small molecule drugs. However, these effects are often not considered in drug development. We aimed to quantitatively evaluate effects of gender and immune status on the PK of PEGylated interferon in frequently used murine models. Chronically cannulated female athymic nude and female and male immunocompetent C57Bl/6J mice (n = 24 in total) received a single intravenous or subcutaneous (s.c.) dose of PEGylated interferon. Serial blood samples were taken for 48 h. Noncompartmental analysis and population PK modeling with covariate analysis were performed to evaluate the data. The PK of PEGylated interferon followed a three compartment disposition model with two sequential compartments for s.c. absorption. Female nude mice had significantly higher plasma clearance than C57Bl/6J mice (0.503 vs. 0.397 mL/h). Male mice had a slower absorption rate constant (0.138 h(-1)) and extent (46.2%) of s.c. absorption than female mice (0.274 in C57Bl/6J and 0.374 h(-1) in nude, 60.8% in both). Thus, gender and immune status significantly impacted on important PK parameters of PEGylated interferon in murine models commonly utilized in drug development. It is critical to take into account these differences when choosing animal models and conducting translational pharmacology research.

PEGylation does not significantly change the initial intravenous or subcutaneous pharmacokinetics or lymphatic exposure of trastuzumab in rats but increases plasma clearance after subcutaneous administration

The lymphatic system plays a major role in the metastatic dissemination of cancer and has an integral role in immunity. PEGylation enhances drainage and lymphatic uptake following subcutaneous (sc) administration of proteins and protein-like polymers, but the impact of PEGylation of very large proteins (such as antibodies) on subcutaneous and lymphatic pharmacokinetics is unknown. This study therefore aimed to evaluate the impact of PEGylation on the sc absorption and lymphatic disposition of the anti-HER2 antibody trastuzumab in rats. PEG-trastuzumab was generated via the conjugation of a single 40 kDa PEG-NHS ester to trastuzumab. PEG-trastuzumab showed a 5-fold reduction in HER2 binding affinity, however the in vitro growth inhibitory effects were preserved as a result of changes in cellular trafficking when compared to native trastuzumab. The lymphatic pharmacokinetics of PEG-trastuzumab was evaluated in thoracic lymph duct cannulated rats after iv and sc administration and compared to the pharmacokinetics of native trastuzumab. The iv pharmacokinetics and lymphatic exposure of PEG-trastuzumab was similar when compared to trastuzumab. After sc administration, initial plasma pharmacokinetics and lymphatic exposure were also similar between PEG-trastuzumab and trastuzumab, but the absolute bioavailability of PEG-trastuzumab was 100% when compared to 86.1% bioavailability for trastuzumab. In contrast to trastuzumab, PEG-trastuzumab showed accelerated plasma clearance beginning approximately 7 days after sc, but not iv, administration, presumably as a result of the generation of anti-PEG IgM. This work suggests that PEGylation does not significantly alter the lymphatic disposition of very large proteins, and further suggests that it is unlikely to benefit therapy with monoclonal antibodies.

Analysis of HGD Gene Mutations in Patients with Alkaptonuria from the United Kingdom: Identification of Novel Mutations

Alkaptonuria (AKU) is a rare autosomal recessive disorder with incidence ranging from 1:100,000 to 1:250,000. The disorder is caused by a deficiency of the enzyme homogentisate 1,2-dioxygenase (HGD), which results from defects in the HGD gene. This enzyme converts homogentisic acid to maleylacetoacetate and has a major role in the catabolism of phenylalanine and tyrosine. To elucidate the mutation spectrum of the HGD gene in patients with alkaptonuria from 42 patients attending the National Alkaptonuria Centre, 14 exons of the HGD gene and the intron-exon boundaries were analysed by PCR-based sequencing. A total of 34 sequence variants was observed, confirming the genetic heterogeneity of AKU. Of these mutations, 26 were missense substitutions and four splice site mutations. There were two deletions and one duplication giving rise to frame shifts and one substitution abolishing the translation termination codon (no stop). Nine of the mutations were previously unreported novel variants. Using computational approaches based on the 3D structure, these novel mutations are predicted to affect the activity of the protein complex through destabilisation of the individual protomer structure or through disruption of protomer-protomer interactions.

Methotrexate-conjugated PEGylated dendrimers show differential patterns of deposition and activity in tumor-burdened lymph nodes after intravenous and subcutaneous administration in rats

The current study sought to explore whether the subcutaneous administration of lymph targeted dendrimers, conjugated with a model chemotherapeutic (methotrexate, MTX), was able to enhance anticancer activity against lymph node metastases. The lymphatic pharmacokinetics and antitumor activity of PEGylated polylysine dendrimers conjugated to MTX [D-MTX(OH)] via a tumor-labile hexapeptide linker was examined in rats and compared to a similar system where MTX was α-carboxyl O-tert-butylated [D-MTX(OtBu)]. The latter has previously been shown to exhibit longer plasma circulation times. D-MTX(OtBu) was well absorbed from the subcutaneous injection site via the lymph, and 3 to 4%/g of the dose was retained by sentinel lymph nodes. In contrast, D-MTX(OH) showed limited absorption from the subcutaneous injection site, but absorption was almost exclusively via the lymph. The retention of D-MTX(OH) by sentinel lymph nodes was also significantly elevated (approximately 30% dose/g). MTX alone was not absorbed into the lymph. All dendrimers displayed lower lymph node targeting after intravenous administration. Despite significant differences in the lymph node retention of D-MTX(OH) and D-MTX(OtBu) after subcutaneous and intravenous administration, the growth of lymph node metastases was similarly inhibited. In contrast, the administration of MTX alone did not significantly reduce lymph node tumor growth. Subcutaneous administration of drug-conjugated dendrimers therefore provides an opportunity to improve drug deposition in downstream tumor-burdened lymph nodes. In this case, however, increased lymph node biodistribution did not correlate well with antitumor activity, possibly suggesting constrained drug release at the site of action.

Crystal structure of human insulin-regulated aminopeptidase with specificity for cyclic peptides

Insulin-regulated aminopeptidase (IRAP or oxytocinase) is a membrane-bound zinc-metallopeptidase that cleaves neuroactive peptides in the brain and produces memory enhancing effects when inhibited. We have determined the crystal structure of human IRAP revealing a closed, four domain arrangement with a large, mostly buried cavity abutting the active site. The structure reveals that the GAMEN exopeptidase loop adopts a very different conformation from other aminopeptidases, thus explaining IRAP's unique specificity for cyclic peptides such as oxytocin and vasopressin. Computational docking of a series of IRAP-specific cognitive enhancers into the crystal structure provides a molecular basis for their structure-activity relationships and demonstrates that the structure will be a powerful tool in the development of new classes of cognitive enhancers for treating a variety of memory disorders such as Alzheimer's disease.

Practical lessons in murine thoracic lymph duct cannulations: observations in female and male mice across four different strains that impact on "cannulatability"

Cannulation of the thoracic lymph duct in experimental animals allows direct measurement of the lymphatic exposure of lymph-targeted drugs. When coupled with recent advances in genetically modified and diseased mouse models, this presents further opportunities to define changes in biological processes and disease in response to drug treatment. Although cannulation of the thoracic lymph duct in mice is inherently challenging because of the small size and delicate nature of the duct, it can be further confounded by anatomical variations between animals. In this communication, we present our observations on the anatomical features of the thoracic lymph duct between mice of different strains and genders, and discuss the impact of these features on the "cannulatability" of the duct. We also provide some technical tips to help guide other investigators to deliver higher experimental success rates.

Platinum: a database of experimentally measured effects of mutations on structurally defined protein-ligand complexes

Drug resistance is a major challenge for the treatment of many diseases and a significant concern throughout the drug development process. The ability to understand and predict the effects of mutations on protein–ligand affinities and their roles in the emergence of resistance would significantly aid treatment and drug design strategies. In order to study and understand the impacts of missense mutations on the interaction of ligands with the proteome, we have developed Platinum (http://structure.bioc.cam.ac.uk/platinum). This manually curated, literature-derived database, comprising over 1000 mutations, associates for the first time experimental information on changes in affinity with three-dimensional structures of protein–ligand complexes. To minimize differences arising from experimental techniques and to directly compare binding affinities, Platinum considers only changes measured by the same group and with the same amino-acid sequence used for structure determination, providing a direct link between protein structure, how a ligand binds and how mutations alter the affinity of the ligand of the protein. We believe Platinum will be an invaluable resource for understanding the effects of mutations that give rise to drug resistance, a major problem emerging in pandemics including those caused by the influenza virus, in infectious diseases such as tuberculosis, in cancer and in many other life-threatening illnesses.

DUET: a server for predicting effects of mutations on protein stability using an integrated computational approach

Cancer genome and other sequencing initiatives are generating extensive data on non-synonymous single nucleotide polymorphisms (nsSNPs) in human and other genomes. In order to understand the impacts of nsSNPs on the structure and function of the proteome, as well as to guide protein engineering, accurate in silicomethodologies are required to study and predict their effects on protein stability. Despite the diversity of available computational methods in the literature, none has proven accurate and dependable on its own under all scenarios where mutation analysis is required. Here we present DUET, a web server for an integrated computational approach to study missense mutations in proteins. DUET consolidates two complementary approaches (mCSM and SDM) in a consensus prediction, obtained by combining the results of the separate methods in an optimized predictor using Support Vector Machines (SVM). We demonstrate that the proposed method improves overall accuracy of the predictions in comparison with either method individually and performs as well as or better than similar methods. The DUET web server is freely and openly available at http://structure.bioc.cam.ac.uk/duet.

Do current therapeutic anti-Aβ antibodies for Alzheimer's disease engage the target?

Reducing amyloid-β peptide (Aβ) burden at the pre-symptomatic stages of Alzheimer's disease (AD) is currently the advocated clinical strategy for treating this disease. The most developed method for targeting Aβ is the use of monoclonal antibodies including bapineuzumab, solanezumab and crenezumab. We have synthesized these antibodies and used surface plasmon resonance (SPR) and mass spectrometry to characterize and compare the ability of these antibodies to target Aβ in transgenic mouse tissue as well as human AD tissue. SPR analysis showed that the antibodies were able to bind Aβ with high affinity. All of the antibodies were able to bind Aβ in mouse tissue. However, significant differences were observed in human brain tissue. While bapineuzumab was able to capture a variety of N-terminally truncated Aβ species, the Aβ detected using solanezumab was barely above detection limits while crenezumab did not detect any Aβ. None of the antibodies were able to detect any Aβ species in human blood. Immunoprecipitation experiments using plasma from AD subjects showed that both solanezumab and crenezumab have extensive cross-reactivity with non-Aβ related proteins. Bapineuzumab demonstrated target engagement with brain Aβ, consistent with published clinical data. Solanezumab and crenezumab did not, most likely as a result of a lack of specificity due to cross-reactivity with other proteins containing epitope overlap. This lack of target engagement raises questions as to whether solanezumab and crenezumab are suitable drug candidates for the preventative clinical trials for AD.

Potent hepatitis C inhibitors bind directly to NS5A and reduce its affinity for RNA

Hepatitis C virus (HCV) infection affects more than 170 million people. The high genetic variability of HCV and the rapid development of drug-resistant strains are driving the urgent search for new direct-acting antiviral agents. A new class of agents has recently been developed that are believed to target the HCV protein NS5A although precisely where they interact and how they affect function is unknown. Here we describe an in vitro assay based on microscale thermophoresis and demonstrate that two clinically relevant inhibitors bind tightly to NS5A domain 1 and inhibit RNA binding. Conversely, RNA binding inhibits compound binding. The compounds bind more weakly to known resistance mutants L31V and Y93H. The compounds do not affect NS5A dimerisation. We propose that current NS5A inhibitors act by favouring a dimeric structure of NS5A that does not bind RNA.

Crystallization and preliminary X-ray diffraction analysis of the Fab portion of the Alzheimer's disease immunotherapy candidate bapineuzumab complexed with amyloid-β

Bapineuzumab (AAB-001) and its derivative (AAB-003) are humanized versions of the anti-Aβ murine antibody 3D6 and are immunotherapy candidates in Alzheimer's disease. The common Fab fragment of these immunotherapies has been expressed, purified and crystallized in complex with β-amyloid peptides (residues 1-8 and 1-28). Diffraction data at high resolution were acquired from crystals of Fab-Aβ8 (2.0 Å) and Fab-Aβ28 (2.2 Å) complexes at the Australian Synchrotron. Both crystal forms belonged to the primitive orthorhombic space group P21221.

Structural studies of Streptococcus pyogenes streptolysin O provide insights into the early steps of membrane penetration

Cholesterol-dependent cytolysins (CDCs) are a large family of bacterial toxins that exhibit a dependence on the presence of membrane cholesterol in forming large pores in cell membranes. Significant changes in the three-dimensional structure of these toxins are necessary to convert the soluble monomeric protein into a membrane pore. We have determined the crystal structure of the archetypical member of the CDC family, streptolysin O (SLO), a virulence factor from Streptococcus pyogenes. The overall fold is similar to previously reported CDC structures, although the C-terminal domain is in a different orientation with respect to the rest of the molecule. Surprisingly, a signature stretch of CDC sequence called the undecapeptide motif, a key region involved in membrane recognition, adopts a very different structure in SLO to that of the well-characterized CDC perfringolysin O (PFO), although the sequences in this region are identical. An analysis reveals that, in PFO, there are complementary interactions between the motif and the rest of domain 4 that are lost in SLO. Molecular dynamics simulations suggest that the loss of a salt bridge in SLO and a cation-pi interaction are determining factors in the extended conformation of the motif, which in turn appears to result in a greater flexibility of the neighboring L1 loop that houses a cholesterol-sensing motif. These differences may explain the differing abilities of SLO and PFO to efficiently penetrate target cell membranes in the first step of toxin insertion into the membrane.

mCSM: predicting the effects of mutations in proteins using graph-based signatures

Motivation: Mutations play fundamental roles in evolution by introducing diversity into genomes. Missense mutations in structural genes may become either selectively advantageous or disadvantageous to the organism by affecting protein stability and/or interfering with interactions between partners. Thus, the ability to predict the impact of mutations on protein stability and interactions is of significant value, particularly in understanding the effects of Mendelian and somatic mutations on the progression of disease. Here, we propose a novel approach to the study of missense mutations, called mCSM, which relies on graph-based signatures. These encode distance patterns between atoms and are used to represent the protein residue environment and to train predictive models. To understand the roles of mutations in disease, we have evaluated their impacts not only on protein stability but also on protein–protein and protein–nucleic acid interactions.

Results: We show that mCSM performs as well as or better than other methods that are used widely. The mCSM signatures were successfully used in different tasks demonstrating that the impact of a mutation can be correlated with the atomic-distance patterns surrounding an amino acid residue. We showed that mCSM can predict stability changes of a wide range of mutations occurring in the tumour suppressor protein p53, demonstrating the applicability of the proposed method in a challenging disease scenario.

Availability and implementation: A web server is available at http://structure.bioc.cam.ac.uk/mcsm.

Contact:dpires@dcc.ufmg.br; tom@cryst.bioc.cam.ac.uk

Supplementary information:Supplementary data are available at Bioinformatics online.