Bioinformatics Services at the OPPF

Oxford Protein Analysis Linker (OPAL)

The Oxford Protein Analysis Linker (OPAL) is a single resource for protein and DNA sequence analysis. It can be used on its own, or more commonly for OPPF users as a front end to the OPTIC database (see below). OPAL incorporates both publicly available analysis tools and bespoke tools developed in-house. These tools include the calculation of parameters from sequence data such as pI, GRAVY, extinction coefficient, rare codon usage, signal sequence and cellular location predictions, potential N- and O-linked glycosylation sites, secondary structure predictions, transmembrane region predictions and potential protease cleavage sites. They also include BLAST-based analyses against publicly available databases including CCD, PDB, SCOP and SWISSPROT. The stand-alone interface to OPAL is publicly available with sequences entered in FASTA format or by GenBank accession number. Results can be displayed visually using the SEView Java applet for browsing molecular sequence data.

OPAL is publicly available. (Note that BLAST against Genbank NR in public version can be slow.)

OPTIC Viewer

The Oxford Protein Target Information Collection (OPTIC) is the main bioinformatics database for storing up-to-date bioinformatics annotations for OPPF targets and serves registered OPPF users. The database management system is SQLServer and not only is OPTIC far faster to use than OPAL (locally licenced versions of software are used), the information is also regularly and automatically refreshed and can generate user alerts. Accession of sequences to the OPTIC database can be trivially scripted and the database has been used for target annotation on the whole-genome scale. The easiest interface to OPTIC is to use OPAL and the SEView applet (see above). An OPPF account is required to use OPTIC.

The OPTIC Viewer is only available to registered users.

PONDR Disorder Prediction Tool

PONDR is perhaps the best-known tool for the prediction of disorder in proteins. We have licenced and installed the VL-XT predictor version of PONDR and it produces a plot very similar that of RONN (see below) which shows predicted probabilities of disorder per residue. PONDR is a neural network method that relies on a parameterization of physico-chemical properties of amino-acids. Since the RONN and PONDR algorithms are very different, it may be valuable to make predictions using both methods and compare the results. PONDR can be run directly from the developers website (www.pondr.com), but using the local version is faster and unlimited.

Incidentally, there are many other methods for prediction of protein disorder available on the web. These include FoldIndex, IUPred, DisEMBL (three versions), GlobPlot, DISOPRED2, DisProt, IST-ZORAN, PreLink and HCA.

Our locally licenced installation of PONDR is only available to local users.

RONN Disorder Prediction Tool

The Regional Order Neural Network (RONN) server was developed as a collaboration between the OPPF and Exeter University as a tool for detecting regions of a sequence which are likely to be disordered. Apart from helping to understand potential functional aspects of a protein sequence, this information is also very valuable in construct design since long disordered sequences can adversely affect protein solubility, crystallizability and diffraction quality. In many cases it may even prevent successful structure determination altogether. The RONN server takes a sequence in FASTA format and produces a plot of the predicted probability of disorder for each residue in the sequence. Residues with a disorder probability >50% are predicted to be disordered. The method involves non-gapped sequence alignment between windows along the query sequence and a set of prototype sequences known to be disordered. A neural network is then used to assign a probability of disorder based on the quality of these alignments. RONN has proven itself to be one of the best methods for disorder prediction and is particularly suited to predictions for partially disordered sequences.

The latest version of RONN is publicly available. The software is also freely distributed to academic institutions.

Target Submission

The Target Submission service is the normal way for registered OPPF users to enter new targets in to Oxford Protein Target Information Collection (OPTIC) bioinformatics database. The system preferably takes the DNA and amino-acid sequences in FASTA format and requests information on organism and the names of responsible and executive OPPF users against whom the target should be registered. When the bioinformatics annotation and database entry is complete the responsible user is notified by email. This service is primarily intended for small numbers of targets. For large numbers of targets, such as complete genomes, please contact the OPPF directly.

The Target Submission service is only available to registered users.