SUMMARY: AlphaPulldown2 streamlines protein structural modeling by automating workflows, improving code adaptability, and optimizing data management for large-scale applications. It introduces an automated Snakemake pipeline, compressed data storage, support for additional modeling backends like UniFold and AlphaLink2, and a range of other improvements. These upgrades make AlphaPulldown2 a versatile platform for predicting both binary interactions and complex multi-unit assemblies. AVAILABILITY AND IMPLEMENTATION: AlphaPulldown2 is freely available at https://github.com/KosinskiLab/AlphaPulldown.

MOTIVATION: Language models are routinely used for text classification and generative tasks. Recently, the same architectures were applied to protein sequences, unlocking powerful new approaches in the bioinformatics field. Protein language models (pLMs) generate high-dimensional embeddings on a per-residue level and encode a "semantic meaning" of each individual amino acid in the context of the full protein sequence. These representations have been used as a starting point for downstream learning tasks and, more recently, for identifying distant homologous relationships between proteins. RESULTS: In this work, we introduce a new method that generates embedding-based protein sequence alignments (EBA) and show how these capture structural similarities even in the twilight zone, outperforming both classical methods as well as other approaches based on pLMs. The method shows excellent accuracy despite the absence of training and parameter optimization. We demonstrate that the combination of pLMs with alignment methods is a valuable approach for the detection of relationships between proteins in the twilight-zone. AVAILABILITY AND IMPLEMENTATION: The code to run EBA and reproduce the analysis described in this article is available at: https://git.scicore.unibas.ch/schwede/EBA and https://git.scicore.unibas.ch/schwede/eba_benchmark.

Today digital revolution is having a dramatic impact on the pharmaceutical industry and the entire healthcare system. The implementation of machine learning, extreme-scale computer simulations, and big data analytics in the drug design and development process offers an excellent opportunity to lower the risk of investment and reduce the time to the patient.

Computational methods for protein structure modelling are routinely used to complement experimental structure determination, thus they help to address a broad spectrum of scientific questions in biomedical research. The most accurate methods today are based on homology modelling, i.e. detecting a homologue to the desired target sequence that can be used as a template for modelling. Here we present a versatile open source homology modelling toolbox as foundation for flexible and computationally efficient modelling workflows. ProMod3 is a fully scriptable software platform that can perform all steps required to generate a protein model by homology. Its modular design aims at fast prototyping of novel algorithms and implementing flexible modelling pipelines. Common modelling tasks, such as loop modelling, sidechain modelling or generating a full protein model by homology, are provided as production ready pipelines, forming the starting point for own developments and enhancements. ProMod3 is the central software component of the widely used SWISS-MODEL web-server.

MOTIVATION: Methods that estimate the quality of a 3D protein structure model in absence of an experimental reference structure are crucial to determine a model's utility and potential applications. Single model methods assess individual models whereas consensus methods require an ensemble of models as input. In this work, we extend the single model composite score QMEAN that employs statistical potentials of mean force and agreement terms by introducing a consensus-based distance constraint (DisCo) score. RESULTS: DisCo exploits distance distributions from experimentally determined protein structures that are homologous to the model being assessed. Feed-forward neural networks are trained to adaptively weigh contributions by the multi-template DisCo score and classical single model QMEAN parameters. The result is the composite score QMEANDisCo, which combines the accuracy of consensus methods with the broad applicability of single model approaches. We also demonstrate that, despite being the de-facto standard for structure prediction benchmarking, CASP models are not the ideal data source to train predictive methods for model quality estimation. For performance assessment, QMEANDisCo is continuously benchmarked within the CAMEO project and participated in CASP13. For both, it ranks among the top performers and excels with low response times. AVAILABILITY AND IMPLEMENTATION: QMEANDisCo is available as web-server at https://swissmodel.expasy.org/qmean. The source code can be downloaded from https://git.scicore.unibas.ch/schwede/QMEAN. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Bioinformatics (2019) doi: 10.1093/bioinformatics/btz828 Correction text: The first version of this article placed the index of summation after, rather than below, the sum symbol in Equations 2, 3, and 4. This has now been corrected online. The publisher apologises for the error.

MOTIVATION: Membrane proteins are an important class of biological macromolecules involved in many cellular key processes including signalling and transport. They account for one third of genes in the human genome and >50% of current drug targets. Despite their importance, experimental structural data are sparse, resulting in high expectations for computational modelling tools to help fill this gap. However, as many empirical methods have been trained on experimental structural data, which is biased towards soluble globular proteins, their accuracy for transmembrane proteins is often limited. RESULTS: We developed a local model quality estimation method for membrane proteins ('QMEANBrane') by combining statistical potentials trained on membrane protein structures with a per-residue weighting scheme. The increasing number of available experimental membrane protein structures allowed us to train membrane-specific statistical potentials that approach statistical saturation. We show that reliable local quality estimation of membrane protein models is possible, thereby extending local quality estimation to these biologically relevant molecules. AVAILABILITY AND IMPLEMENTATION: Source code and datasets are available on request. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: The assessment of protein structure prediction techniques requires objective criteria to measure the similarity between a computational model and the experimentally determined reference structure. Conventional similarity measures based on a global superposition of carbon α atoms are strongly influenced by domain motions and do not assess the accuracy of local atomic details in the model. RESULTS: The Local Distance Difference Test (lDDT) is a superposition-free score that evaluates local distance differences of all atoms in a model, including validation of stereochemical plausibility. The reference can be a single structure, or an ensemble of equivalent structures. We demonstrate that lDDT is well suited to assess local model quality, even in the presence of domain movements, while maintaining good correlation with global measures. These properties make lDDT a robust tool for the automated assessment of structure prediction servers without manual intervention. AVAILABILITY AND IMPLEMENTATION: Source code, binaries for Linux and MacOSX, and an interactive web server are available at http://swissmodel.expasy.org/lddt. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: MODalign is an interactive web-based tool aimed at helping protein structure modelers to inspect and manually modify the alignment between the sequences of a target protein and of its template(s). It interactively computes, displays and, upon modification of the target-template alignment, updates the multiple sequence alignments of the two protein families, their conservation score, secondary structure and solvent accessibility values, and local quality scores of the implied three-dimensional model(s). Although it has been designed to simplify the target-template alignment step in modeling, it is suitable for all cases where a sequence alignment needs to be inspected in the context of other biological information. AVAILABILITY AND IMPLEMENTATION: Freely available on the web at http://modorama.biocomputing.it/modalign. Website implemented in HTML and JavaScript with all major browsers supported. CONTACT: [email protected].

This special issue comprises the proceedings papers accepted for presentation at the 11th European Conference on Computational Biology (ECCB 2012), which was held from 9 to 12 September 2012 at the Conference Center Basel, Switzerland. This year's conference was organized by the SIB Swiss Institute of Bioinformatics and also served as the 10th [BC]2 Basel Computational Biology Conference, the annual meeting of the Swiss computational biology community. It was preceded by two satellite meetings, six workshops and five tutorials. Details of the conference are available through the conference website (http://www.eccb12.org) and will later be archived at eccb.iscb.org/2012/. ECCB is the top European conference in computational biology and bioinformatics, and together with ISMB and RECOMB one of the major international conference series in computational biology. The conference brings together scientists working in a broad range of disciplines, including bioinformatics, computational and molecular biology, medicine and systems biology. It covers computational methods for analysis of the tremendously growing amount of data in the life sciences, and approaches for modeling and simulation of increasingly complex systems in molecular biology, medicine and pharmaceutical research. The scope of the conference evolves each year to address the latest developments in biological applications of mathematical modeling and computational methods. ECCB is held in a different country or region each year, and incorporates the annual national or regional meeting where it is held. Every other year, it is held jointly with the Intelligent Systems in Molecular Biology (ISMB). Past editions of ECCB have been held in Vienna (AT) (with ISMB), Ghent (BE), Stockholm (SE) (joint with ISMB), Cagliari (IT), Vienna (AT) (with ISMB), Eilat (IL), Madrid (ES), Glasgow (UK) (with ISMB), Paris (FR) and Saarbrücken (DE). The ECCB 2012 conference featured keynote lectures by distinguished speakers: The opening keynote lecture was delivered by Nobel price laureate Aaron Ciechanover (Technion - Israel Institute of Technology), followed by Barry Honig (Columbia University, New York), Laurent Keller (University of Lausanne), Gene Myers (MPI of Molecular Cell Biology and Genetics, Dresden), Mihaela Zavolan (Biozentrum University of Basel & SIB) and the EMBO keynote lecture by Søren Brunak (Technical University of Denmark). The conference topics span all areas of methodological developments for computational biology and innovative applications of computational methods to molecular biology. This year, the conference was organized in 11 topic areas: ‘Applied and Translational Bioinformatics’, ‘Bioimaging, Spatial-temporal Modeling and Data Visualization’, ‘Bioinformatics of Health and Disease, Biomarkers and Personalized Medicine’, ‘Databases, Ontologies, and Text Mining’, ‘Evolution, Phylogeny, and Comparative Genomics’, ‘Macromolecular Structure, Dynamics and Function’, ‘Mutations, Variations, and Population Genomics’, ‘Protein Interactions, Molecular Networks, and Proteomics’, ‘Regulation, Pathways, and Systems Biology’, ‘Sequencing and Sequence Analysis’ and ‘Technology and Software (TechTrack)’. Following the call for proceedings papers, we received a record number of 340 submissions—which corresponds to a ~60% increase compared to the previous ECCB in 2010. The manuscripts were assessed by a program committee of 350 reviewers and 191 co-reviewers. The reviewing process in each topic area was coordinated by two area chairs (see next article), and the different areas were coordinated by the Conference Chairs. Within each area, the area chairs assigned the papers to expert referees; in general each paper was reviewed by three PC members, with a few cases where we could only secure two referee reports. The whole process was carried out through the EasyChair conference reviewing system (http://www.easychair.org). The same system was used for discussing papers in cases where the reviewers had different opinions about the relevance or quality of the contribution. After a consensus had been reached by the reviewers, a final selection was carried out by the area chairs and the conference chairs, discussing the outcome in the form of a phone conference. In total, 48 papers were conditionally accepted, pending revision. After acceptance notices were sent, the authors had 2 weeks to modify their papers according to the suggestions made by the reviewers and to respond to the reviewers’ comments. Modified papers and authors’ responses were reexamined during the next week to ensure each paper was modified appropriately in response to the reviewers’ comments. We were pleased that all conditionally accepted papers were finally accepted. These 48 (14% accepted rate) contributions are included in this special issue. Collectively, the authors of the 48 accepted papers come from 19 countries across 4 continents, with 60 authors from North America, 95 from Europe, 27 from Asia (excluding Israel), 4 from Israel and 3 authors from Saudi Arabia. The conference papers are available in open access in electronic format, both through online open access to the journal Bioinformatics and on the electronic documentation distributed to the conference participants on USB memory cards. The call for poster abstract submission was kept open 2 weeks after the selection of the oral presentations, so that authors whose manuscript could not be accepted had the opportunity to resubmit their work as poster. A call for late poster abstracts was launched to allow groups to present their latest research results during the meeting. At the time of writing, more than 500 posters were accepted after evaluation of the abstracts for scientific relevance to the meeting. The call for late abstract has not been completed at the time of writing. Accepted poster abstracts are included as part of the electronic documentation distributed to the conference participants, and are available through the conference website. In addition to the keynotes and proceedings presentations, eight highlight presentations were selected by the ECCB steering board and program committee members from 100 submissions, consisting of full papers that have been published, or accepted for publication, between 1 January 2011 and the submission deadline, 1 July 2012. Additionally, a technology and software track included presentations of commercial and academic software, databases and biological applications by sponsors, exhibitors and academic institutions. Sixteen exhibitor stands were open throughout the conference in the main conference halls, presenting the latest scientific literature in the field of computational biology, bioinformatics, modeling and simulation, as well as new hardware, software and technology developments. ECCB 2012 also hosted the first technical meeting of ELIXIR chaired by Søren Brunak (chair of the ELIXIR Interim Board). During the weekend preceding the conference, two satellite meetings, six workshops and five tutorials were organized: on Saturday, 8 September, a special symposium celebrated the 10th anniversary of the UniProt project: in 2002, 40 years after the first comprehensive collection of protein sequences, the ‘Atlas of Protein Sequence and Structure’, was published, the SWISS-PROT and TrEMBL protein sequence databases at the European Bioinformatics Institute (EBI) and the SIB Swiss Institute of Bioinformatics, and the Protein Information Resource (PIR) in Washington, joined forces to create UniProt. At this symposium, renowned speakers gave insight into their work, highlighting how protein databases are underpinning life sciences. The Student Council of the International Society for Computational Biology organized its second European Student Council Symposium (ESCS), chaired by Pedro Lopes (University of Aveiro, Portugal) and Florian Heer (SIB & Biozentrum University of Basel, Switzerland). The meeting gave students the opportunity to present their work to an international audience and to build a network and exchange ideas and knowledge within the computational biology community. On Saturday and Sunday, 8 and 9 September, the Sixth International Workshop on Machine Learning in Systems Biology (MLSB) was organized as a 2-day workshop by Karsten Borgwardt (Max Planck Institutes & Eberhard Karls Universität Tübingen) and Gunnar Rätsch (Sloan-Kettering Institute, New York). MLSB provides a scientific forum for the exchange between researchers from Systems Biology and Machine Learning. All other workshops and tutorials were organized as 1-day events on Sunday 9 September: the workshop ‘Annotation, Interpretation and Management of Mutations (AIMM)’ focused on extraction and reuse of genotype–phenotype knowledge from scientific literature and databases. It was chaired by Christopher Baker (University of New Brunswick, Saint John, Canada) and Dietrich Rebholz-Schuhmann (European Bioinformatics Institute, Cambridge, UK). The workshop ‘Computational Proteomics: From Mass Spectrometry to Protein Complexes’ organized by Markus Müller (SIB, Geneva), Alexander Schmidt (Biozentrum, University of Basel) and Jacques Colinge (CeMM, Vienna) addressed computational aspects of Mass spectrometry analysis. The workshop ‘Bioinformatics Training for Life Scientists: Showcases and Challenges from Tutors’ Perspectives’ organized by Patricia Palagi (SIB Swiss Institute of Bioinformatics), Vicky Schneider (European Bioinformatics Institute, EBI), Allegra Via (Dept of Physics, Sapienza University of Rome), Celia van Gelder (Netherlands Bioinformatics Centre, NBIC) focused on training in bioinformatics for life scientists from the perspective of the trainer. It aimed to provide a platform for those involved in bioinformatics training to meet, confront and share experiences, as well as discuss future directions and collaboration in joint efforts. The workshop ‘Imaging Analysis and Computational Modeling for Cancer and its Therapy’ organized by Petros Koumoutsakos and Gabor Szekely (ETHZ) presented the state of the art in imaging, analysis and modeling of cancer and its therapy. Special focus was laid on recent developments in optical microscopy, small animal and clinical imaging, data analysis tools, mathematical models and computations of cancer formation and progression. The workshop ‘Detecting Transcription Factor Binding Sites with ChIP-Seq Data and Predicting Damaging Cis-Regulatory Variations’ addressed genome re-sequencing data analysis for the prediction of variations within cis-regulatory sequences. The program was organized by Wyeth W. Wasserman (University of British Columbia, Canada), Virginie Bernard (Curie Institute, Canada), Jonathan Lim and Anthony Mathelier (University of British Columbia, Canada). The purpose of the tutorial program at ECCB is to provide participants with lectures and hands-on training covering topics relevant to the bioinformatics field. It offers participants an opportunity to learn about new areas of bioinformatics research, to get an introduction to important established topics, or to develop advanced skills in areas they are already familiar with. At ECCB 2012, tutorials on six different topics were organized: ‘Applications of Bio-Ontologies in Large-Scale Data-Driven Science: A Practical Introduction’ by Barry Smith (University at Buffalo, USA), and Janna Hastings (European Bioinformatics Institute and University of Geneva); ‘Protein Evolution: From Sequence to Structure to Function’ by Christine Orengo and Romain Studer (UCL, London, UK) and Nicholas Furnham (EMBL-EBI, Cambridge, UK); ‘GWAS: Statistics and Bioinformatics to Analyse Genotype to Phenotype Relationships from Molecular Traits to Disease Phenotypes’ by Christoph Lippert (Microsoft Research, Los Angeles, USA) and Oliver Stegle (Max Planck Institute for Developmental Biology, Tübingen, Germany); ‘Inferring Genetic Diversity from Next-generation Sequencing Data: Computational Methods and Biomedical Applications’ by Niko Beerenwinkel (ETH Zurich and SIB), Karin J. Metzner (University Hospital Zurich) and Volker Roth (University of Basel, Switzerland); and ‘Reads to Biological Patterns: End-to-End Differential Expression Analysis of RNA Sequencing Data Using Bioconductor’ by Simon Anders and Wolfgang Huber (EMBL), Michal Okoniewski (ETHZ) and Mark Robinson (University of Zurich & SIB). Encouraging the participation of young scientists has always been a central goal of the ECCB conference because the conference is a great opportunity for education and networking for early stage scientists. Thanks to contributions from several academic sponsors, our host organizations, previous editions of ECCB and our commercial sponsors, we could provide 114 conference fellowships to PhD and postdoc applicants from all over the world. Presenters and contributors to oral presentations were given priority. Unfortunately, several applicants had to be turned down due to limitations of available resources. Organizing ECCB would not be possible without the generous support by partner organizations, funding agencies and sponsors. We are especially grateful to the Swiss Foundation for Excellence and Talent in Biomedical Research, SIB Swiss Institute of Bioinformatics, ISCB International Society for Computational Biology, the University of Basel, the Swiss Academy of Medical Sciences and the [BC]2 Basel Computational Biology Conference for funding 114 fellowships to facilitate the participation of young talented scientists early in their career. We highly appreciate the support by our host institutions, the SIB Swiss Institute of Bioinformatics, the Biozentrum University of Basel, and the Swiss Systems Biology Initiative SystemsX.ch. We are especially grateful to our platinum sponsors Basel Area and IBM. We would like to thank all our partner and sponsors: the Netherlands Bioinformatics Centre NBIC, CONVEY Computers, F. Hoffmann-La Roche Ltd, HP, BBRC, Interpharma, FMI Friedrich Miescher Institute, Bioalps, Competence Center Computational Sciences University Basel, Novartis Pharma, Syngenta, BASF, Merck Serono and the canton of the city of Basel. We would like to thank EMBO for supporting the EMBO keynote lecture by Soren Brunak. It is our pleasure to thank all the people and organizations that made ECCB 2012 possible. Our first thanks go to the program committee of reviewers and co-reviewers, and in particular the area chairs. A swift and effective review process is at the heart of a conference like ECCB. Reviewing 340 proceedings manuscripts and 100 highlight paper submissions in just a few weeks is truly challenging, especially since many members of the program committee had been involved in reviewing for ISMB shortly before. The dedication and hard work of the program committee was quite simply the key to the high quality of the conference. We are grateful to the ECCB steering committee for their support to the organization of the conference: in particular, the advice and guidance from previous organizers Anna Tramontano, Yves Moreau and Michal Linial was crucial. Anna, thank you so much for answering our questions at any time of day (and night)! Logistics is crucial for a successful conference. We are grateful to the Oxford University Press production team for the preparation of this special issue, to Congress Center Basel for the venue organization. The local administration team at the Biozentrum of the University of Basel did an incredible job, and we owe them a great debt: Katja Jenni, Rita Manohar, Sarah Güthe, Lorenza Bordoli and Yvonne Steger for preparing the meeting and organizing the registration, Markus Stöckli from Billboard for the poster and website design, Andrew Waterhouse for the registration and poster submission system, JanWelker and Konstantin Arnold for IT support, Jocelyne Boquet for handling the financial administration. Finally, the most important ingredient to the success of a conference are the scientific contributions and discussions: we are grateful to the authors of all submitted manuscripts, keynote speakers, presenters of selected papers, tutorial lecturers, workshop organizers and contributors, poster authors and to all the participants. Thank you for making this conference a success.

MOTIVATION: Quality assessment of protein structures is an important part of experimental structure validation and plays a crucial role in protein structure prediction, where the predicted models may contain substantial errors. Most current scoring functions are primarily designed to rank alternative models of the same sequence supporting model selection, whereas the prediction of the absolute quality of an individual protein model has received little attention in the field. However, reliable absolute quality estimates are crucial to assess the suitability of a model for specific biomedical applications. RESULTS: In this work, we present a new absolute measure for the quality of protein models, which provides an estimate of the 'degree of nativeness' of the structural features observed in a model and describes the likelihood that a given model is of comparable quality to experimental structures. Model quality estimates based on the QMEAN scoring function were normalized with respect to the number of interactions. The resulting scoring function is independent of the size of the protein and may therefore be used to assess both monomers and entire oligomeric assemblies. Model quality scores for individual models are then expressed as 'Z-scores' in comparison to scores obtained for high-resolution crystal structures. We demonstrate the ability of the newly introduced QMEAN Z-score to detect experimentally solved protein structures containing significant errors, as well as to evaluate theoretical protein models. In a comprehensive QMEAN Z-score analysis of all experimental structures in the PDB, membrane proteins accumulate on one side of the score spectrum and thermostable proteins on the other. Proteins from the thermophilic organism Thermatoga maritima received significantly higher QMEAN Z-scores in a pairwise comparison with their homologous mesophilic counterparts, underlining the significance of the QMEAN Z-score as an estimate of protein stability. AVAILABILITY: The Z-score calculation has been integrated in the QMEAN server available at: http://swissmodel.expasy.org/qmean.

MOTIVATION: Developers of new methods in computational structural biology are often hampered in their research by incompatible software tools and non-standardized data formats. To address this problem, we have developed OpenStructure as a modular open source platform to provide a powerful, yet flexible general working environment for structural bioinformatics. OpenStructure consists primarily of a set of libraries written in C++ with a cleanly designed application programmer interface. All functionality can be accessed directly in C++ or in a Python layer, meeting both the requirements for high efficiency and ease of use. Powerful selection queries and the notion of entity views to represent these selections greatly facilitate the development and implementation of algorithms on structural data. The modular integration of computational core methods with powerful visualization tools makes OpenStructure an ideal working and development environment. Several applications, such as the latest versions of IPLT and QMean, have been implemented based on OpenStructure-demonstrating its value for the development of next-generation structural biology algorithms. AVAILABILITY: Source code licensed under the GNU lesser general public license and binaries for MacOS X, Linux and Windows are available for download at http://www.openstructure.org. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Tandem mass spectrometry is a powerful technique for the identification of proteins contained in a biological sample. We present a grid-based application service for the analysis of tandem mass spectra, based on the open-source X/Tandem algorithm. The implemented parallelization strategy allows the distribution of analysis jobs to a grid of desktop PCs, utilizing their idle capacity. For this application service, a user-friendly project-oriented frontend was developed, implemented as a Web-based workspace. After analysis, results are visualized and presented to the user along with a set of powerful and novel tools to allow further interpretation and evaluation of search results.

Modern biology has become a science of information, analysis and prediction, coalescing into computational biology -- a single discipline at the crossroads of life sciences, informatics, and mathematics. New developments in information and communications technology as well as high-performance computing enable researchers to address new demanding scientific problems which seemed far out of reach only a few years ago. The computational requirements of most applications in computational biology differ significantly from the requirements of other users of highthroughput computing such as high energy physics. To address these needs, the SwissBioGrid initiative, a collaboration among several partner institutions with a broad spectrum of expertise, was started over a year ago. In this paper, we report on its current status and achievements as well as the lessons learned which are of interest to the wider e-Science and Grid communities.

MOTIVATION: Homology models of proteins are of great interest for planning and analysing biological experiments when no experimental three-dimensional structures are available. Building homology models requires specialized programs and up-to-date sequence and structural databases. Integrating all required tools, programs and databases into a single web-based workspace facilitates access to homology modelling from a computer with web connection without the need of downloading and installing large program packages and databases. RESULTS: SWISS-MODEL workspace is a web-based integrated service dedicated to protein structure homology modelling. It assists and guides the user in building protein homology models at different levels of complexity. A personal working environment is provided for each user where several modelling projects can be carried out in parallel. Protein sequence and structure databases necessary for modelling are accessible from the workspace and are updated in regular intervals. Tools for template selection, model building and structure quality evaluation can be invoked from within the workspace. Workflow and usage of the workspace are illustrated by modelling human Cyclin A1 and human Transmembrane Protease 3. AVAILABILITY: The SWISS-MODEL workspace can be accessed freely at http://swissmodel.expasy.org/workspace/