A hybrid clustering of protein binding sites

SCARF: a biomedical association rule finding webserver

The analysis of enormous datasets with missing data entries is a standard task in biological and medical data processing. Large-scale, multi-institution clinical studies are the typical examples of such datasets. These sets make possible the search for multi-parametric relations since from the plenty of the data one is likely to find a satisfying number of subjects with the required parameter ensembles. Specifically, finding combinatorial biomarkers for some given condition also needs a very large dataset to analyze. For fast and automatic multi-parametric relation discovery association-rule finding tools are used for more than two decades in the data-mining community. Here we present the SCARF webserver for generalized association rule mining. Association rules are of the form: a AND b AND … AND x → y, meaning that the presence of properties a AND b AND … AND x implies property y; our algorithm finds generalized association rules, since it also finds logical disjunctions (i.e., ORs) at the left-hand side, allowing the discovery of more complex rules in a more compressed form in the database. This feature also helps reducing the typically very large result-tables of such studies, since allowing ORs in the left-hand side of a single rule could include dozens of classical rules. The capabilities of the SCARF algorithm were demonstrated in mining the Alzheimer’s database of the Coalition Against Major Diseases (CAMD) in our recent publication (Archives of Gerontology and Geriatrics Vol. 73, pp. 300–307, 2017). Here we describe the webserver implementation of the algorithm.

PDB_Amyloid: an extended live amyloid structure list from the PDB

The Protein Data Bank (PDB) contains more than 135 000 entries at present. From these, relatively few amyloid structures can be identified, since amyloids are insoluble in water. Therefore, most amyloid structures deposited in the PDB are in the form of solid state NMR data. Based on the geometric analysis of these deposited structures, we have prepared an automatically updated web server, which generates a list of the deposited amyloid structures, and also entries of globular proteins that have amyloid‐like substructures of given size and characteristics. We have found that by applying only appropriately selected geometric conditions, it is possible to identify deposited amyloid structures and a number of globular proteins with amyloid‐like substructures. We have analyzed these globular proteins and have found proof in the literature that many of them form amyloids more easily than many other globular proteins. Our results relate to the method of Stanković et al. [Stanković I et al. (2017) IPSI BgD Tran Int Res 13, 47–51], who applied a hybrid textual‐search and geometric approach for finding amyloids in the PDB. If one intends to identify a subset of the PDB for certain applications, the identification algorithm needs to be re‐run periodically, since in 2017 on average 30 new entries per day were deposited in the data bank. Our web server is updated regularly and automatically, and the identified amyloid and partial amyloid structures can be viewed or their list can be downloaded from the following website https://pitgroup.org/amyloid.

Sequence-based prediction of physicochemical interactions at protein functional sites using a function-and-interaction-annotated domain profile database

Background

Identifying protein functional sites (PFSs) and, particularly, the physicochemical interactions at these sites is critical to understanding protein functions and the biochemical reactions involved. Several knowledge-based methods have been developed for the prediction of PFSs; however, accurate methods for predicting the physicochemical interactions associated with PFSs are still lacking.

Results

In this paper, we present a sequence-based method for the prediction of physicochemical interactions at PFSs. The method is based on a functional site and physicochemical interaction-annotated domain profile database, called fiDPD, which was built using protein domains found in the Protein Data Bank. This method was applied to 13 target proteins from the very recent Critical Assessment of Structure Prediction (CASP10/11), and our calculations gave a Matthews correlation coefficient (MCC) value of 0.66 for PFS prediction and an 80% recall in the prediction of the associated physicochemical interactions.

Conclusions

Our results show that, in addition to the PFSs, the physical interactions at these sites are also conserved in the evolution of proteins. This work provides a valuable sequence-based tool for rational drug design and side-effect assessment. The method is freely available and can be accessed at http://202.119.249.49.

Mapping the 3D structures of small molecule binding sites

Background

Analysis of the 3D structures of protein–ligand binding sites can provide valuable insights for drug discovery. Binding site comparison (BSC) studies can be employed to elucidate the function of orphan proteins or to predict the potential for polypharmacology. Many previous binding site analyses only consider binding sites surrounding an experimentally observed bound ligand.

Results

To encompass potential protein–ligand binding sites that do not have ligands known to bind, we have incorporated fpocket cavity detection software and assessed the impact of this inclusion on BSC performance. Using fpocket, we generated a database of ligand-independent potential binding sites and applied the BSC tool, SiteHopper, to analyze similarity relationships between protein binding sites. We developed a method for clustering potential binding sites using a curated dataset of structures for six therapeutically relevant proteins from diverse protein classes in the protein data bank. Two clustering methods were explored; hierarchical clustering and a density-based method adept at excluding noise and outliers from a dataset. We introduce circular plots to visualize binding site structure space. From the datasets analyzed in this study, we highlight a structural relationship between binding sites of cationic trypsin and prothrombin, protein targets known to bind structurally similar small molecules, exemplifying the potential utility of objectively and holistically mapping binding site space from the structural proteome.

Conclusions

We present a workflow for the objective mapping of potential protein–ligand binding sites derived from the currently available structural proteome. We show that ligand-independent binding site detection tools can be introduced without excessive penalty on BSC performance. Clustering combined with intuitive visualization tools can be applied to map relationships between the 3D structures of protein binding sites.Graphical abstract

A PDB-wide, evolution-based assessment of protein-protein interfaces

Background

Thanks to the growth in sequence and structure databases, more than 50 million sequences are now available in UniProt and 100,000 structures in the PDB. Rich information about protein–protein interfaces can be obtained by a comprehensive study of protein contacts in the PDB, their sequence conservation and geometric features.

Results

An automated computational pipeline was developed to run our Evolutionary Protein–Protein Interface Classifier (EPPIC) software on the entire PDB and store the results in a relational database, currently containing > 800,000 interfaces. This allows the analysis of interface data on a PDB-wide scale. Two large benchmark datasets of biological interfaces and crystal contacts, each containing about 3000 entries, were automatically generated based on criteria thought to be strong indicators of interface type. The BioMany set of biological interfaces includes NMR dimers solved as crystal structures and interfaces that are preserved across diverse crystal forms, as catalogued by the Protein Common Interface Database (ProtCID) from Xu and Dunbrack. The second dataset, XtalMany, is derived from interfaces that would lead to infinite assemblies and are therefore crystal contacts. BioMany and XtalMany were used to benchmark the EPPIC approach. The performance of EPPIC was also compared to classifications from the Protein Interfaces, Surfaces, and Assemblies (PISA) program on a PDB-wide scale, finding that the two approaches give the same call in about 88% of PDB interfaces. By comparing our safest predictions to the PDB author annotations, we provide a lower-bound estimate of the error rate of biological unit annotations in the PDB. Additionally, we developed a PyMOL plugin for direct download and easy visualization of EPPIC interfaces for any PDB entry. Both the datasets and the PyMOL plugin are available at http://www.eppic-web.org/ewui/#downloads.

Conclusions

Our computational pipeline allows us to analyze protein–protein contacts and their sequence conservation across the entire PDB. Two new benchmark datasets are provided, which are over an order of magnitude larger than existing manually curated ones. These tools enable the comprehensive study of several aspects of protein–protein contacts in the PDB and represent a basis for future, even larger scale studies of protein–protein interactions.