- Open Access
MeDor: a metaserver for predicting protein disorder
BMC Genomicsvolume 9, Article number: S25 (2008)
We have previously shown that using multiple prediction methods improves the accuracy of disorder predictions. It is, however, a time-consuming procedure, since individual outputs of multiple predictions have to be retrieved, compared to each other and a comprehensive view of the results can only be obtained through a manual, fastidious, non-automated procedure. We herein describe a new web metaserver, MeDor, which allows fast, simultaneous analysis of a query sequence by multiple predictors and provides a graphical interface with a unified view of the outputs.
MeDor was developed in Java and is freely available and downloadable at: http://www.vazymolo.org/MeDor/index.html. Presently, MeDor provides a HCA plot and runs a secondary structure prediction, a prediction of signal peptides and transmembrane regions and a set of disorder predictions. MeDor also enables the user to customize the output and to retrieve the sequence of specific regions of interest.
As MeDor outputs can be printed, saved, commented and modified further on, this offers a dynamic support for the analysis of protein sequences that is instrumental for delineating domains amenable to structural and functional studies.
In recent years there has been an increasing amount of experimental evidence pointing out the abundance of protein disorder within the living world. Recent computational studies have shown that the frequency and length of disordered regions increases with increasing organism complexity (see [1, 2]), with as much as one third of eukaryotic proteins containing long intrinsically disordered regions  and 12% of them being fully disordered .
The identification of disordered regions has a practical interest. Disordered regions often have a biased amino acid composition that can lead to spurious sequence similarity with unrelated proteins. The recognition of regions of disorder is thus crucial to avoid spurious sequence alignments with sequences of unrelated, structured proteins. Secondly, the recognition of disordered regions facilitates the identification of eukaryotic linear motifs, which are short functional motifs occurring mainly (>70%) within disordered regions [5–8], and the functional annotation of proteins . Last, but not least, disordered regions often prevent crystallization of proteins, or the generation of interpretable NMR data and thus represent a bottleneck in high throughput structural determination . As such, their identification is instrumental for delineating protein domains amenable to crystallization and/or to dissect target sequences into a set of independently folded domains in order to facilitate tertiary structure and threading predictions (see [11, 12]).
Intrinsically disordered proteins possess distinctive sequence features, including paucity of hydrophobic residues and enrichment in hydrophilic residues, which allow them to be predicted with a rather good accuracy. Based on these peculiar sequence properties, a series of predictors have been developed in the last years, the majority of which are available on the web (for reviews see [13–16]). As different "flavors" of disorder exist , a given predictor may be more performant in detecting a given "flavor" of disorder against which it has been trained. Hence, the reliability of disorder prediction benefits from the use of several methods relying on different concepts or different physico-chemical parameters. Indeed, we have previously shown that predictions good enough to decipher the modular organization of a protein can only be obtained by combining various predictors (for examples see [11, 14, 15, 18–21]). However this is very time-consuming, since multiple predictions have to be carried out, individual outputs have to be retrieved, compared to each other and a comprehensive view of the results can only be obtained through a manual, fastidious, non-automated procedure. This prompted us to develop MeDor (MEtaserver of DisORder), which is a tool for speeding up the analysis of protein disorder thanks to the simultaneous submission and retrieval of multiple disorder predictions.
MeDor is a Java2 application that provides a graphical output summarizing the predictions of the following programs: a secondary structure prediction (SSP), based on the StrBioLib library of the Pred2ary program [22, 23], Hydrophobic Cluster Analysis (HCA) , IUPred , Prelink , RONN , FoldUnfold , DisEMBL , FoldIndex , GlobPlot2 , DISPROT VL3 and VL3H , DISPROT VSL2B  and Phobius . SSP and HCA have been included in the MeDor program and do not require a web connection. For predictors remotely launched through connection to the public web servers, we selected predictors that (i) rely on different physico-chemical principles, (ii) return results online with a delay compatible with the interactive character of the tool, and (ii) do not require an e-mail address. Additional predictors could be nevertheless easily implemented in MeDor in the future. Of the three predictions provided by DisEMBL, only Rem 465 provides disorder predictions  and is therefore run by default. The two other DisEMBL predictions, i.e. "Loops" and "Hot Loops", can be optionally selected from the MeDor input frame. These latter are indeed useful in terms of identification of regions devoid of regular secondary structure. As such, they are complementary to SSP, with the "Hot Loops" prediction providing in addition information on the extent of thermal agitation.
All requests are submitted in parallel by launching multiple predictors and using default parameters. Retrieval of results is fast, as it takes at maximum the time required by the slowest predictor to reply (in the limit of the timeout fixed by the user) plus the connection time to the server (limited to 15 seconds). SSP in MeDor is run using the medium jury of Pred2ary, which provides a good compromise between accuracy and rapidity. HCA makes use of a two-dimensional helical representation of protein sequences, and thus it is not stricto sensu a predictor. In HCA plots, disordered regions are recognizable as they are depleted in hydrophobic clusters .
For predictors that provide boundaries between ordered and disordered regions, these latter are directly extracted from the outputs. For predictors that provide only disorder probabilities, MeDor applies a 50% cutoff to assign disorder. Beyond disorder predictors, the Phobius program , which predicts transmembrane domains and signal peptides, can also be optionally selected from the MeDor input frame.
The input format is a single protein amino acid sequence in either plain text or FASTA. The input window allows selection of the predictors to be run and choice of the timeout, which can be set in the range of 1 second to 30 minutes.
MeDor provides a graphical output (Fig. 1), in which the sequence query and the results of the various predictors are featured horizontally, with a scroll bar allowing progression from the N-terminus to the C-terminus. All predictions are drawn along the sequence that is represented as a single, continuous horizontal line. Whenever provided by the disorder predictors, per residue probabilities are included in the MeDor output and shown in the status bar.
The graphical output is a dynamic interface that can be customized. It allows the user to display a graphical repository line for comparison among different predictions, to highlight zones of interest, to retrieve sequence boundaries for each prediction and to extract parts of sequences corresponding to a prediction or a highlighted area. The main menu of the graphical output gives access to several functions, such as focus in and out, display/hide the left description labels or the results of any of the predictors, select the highlight color, and display/hide the comment panel. This latter allows insertion of a text/comment in the rich text format (rtf). This main menu also contains options for printing the graphic output, for saving it either as a png image or as an XLM-based file format specific to MeDor (.med). The «Load» function from the input window menu of MeDor allows the user to load a file in the ".med" format. The whole MeDor functionalities are described in the associated help file.
MeDor is not intended to provide a consensus of disorder prediction. Rather, it is conceived to provide a global overview of various predictions relying on different philosophies, and to speed up the disorder prediction step by itself. In addition to the fast identification of regions of disorder, MeDor can also be used to infer information on secondary structure elements and on the possible occurrence of transmembrane regions and signal peptides. Future developments of MeDor may involve generation of a consensus of disorder prediction, which is expected to further accelerate the process of deciphering the modular organization of proteins. Finally, as MeDor outputs can be saved, commented and modified further on, this offers a dynamic support for the analysis of protein sequences that is expected to be very useful in the context of collaborative projects involving several partners. As such, MeDor will facilitate the definition of domain boundaries amenable to structural and functional studies within proteins targeted by structural genomics consortia, such as VIZIER http://www.vizier-europe.org.
Availability and requirements
Project name: MeDor
MeDor home page: http://www.vazymolo.org/MeDor/index.html
Operating systems: Platform independent
Programming language: Java
Other requirements: Java 1.5.0 or higher and a web connection
License: This program uses predictions incoming from public web-servers and is provided freely and "as it is" without any warranty of any kind, either expressed or implied.
Any restrictions to use by non-academics: none.
Dunker AK, Obradovic Z: The protein trinity – linking function and disorder. Nat Biotechnol. 2001, 19 (9): 805-806.
Oldfield CJ, Cheng Y, Cortese MS, Brown CJ, Uversky VN, Dunker AK: Comparing and Combining Predictors of Mostly Disordered Proteins. Biochemistry. 2005, 44 (6): 1989-2000.
Ward JJ, Sodhi JS, McGuffin LJ, Buxton BF, Jones DT: Prediction and functional analysis of native disorder in proteins from the three kingdoms of life. J Mol Biol. 2004, 337 (3): 635-645.
Bogatyreva NS, Finkelstein AV, Galzitskaya OV: Trend of amino acid composition of proteins of different taxa. J Bioinform Comput Biol. 2006, 4 (2): 597-608.
Iakoucheva LM, Brown CJ, Lawson JD, Obradovic Z, Dunker AK: Intrinsic disorder in cell-signaling and cancer-associated proteins. J Mol Biol. 2002, 323 (3): 573-584.
Puntervoll P, Linding R, Gemund C, Chabanis-Davidson S, Mattingsdal M, Cameron S, Martin DM, Ausiello G, Brannetti B, Costantini A: ELM server: A new resource for investigating short functional sites in modular eukaryotic proteins. Nucleic Acids Res. 2003, 31 (13): 3625-3630.
Linding R: Linear Functional Modules. Implication for protein function. Heidelberg.
Neduva V, Linding R, Su-Angrand I, Stark A, Masi FD, Gibson TJ, Lewis J, Serrano L, Russell RB: Systematic Discovery of New Recognition Peptides Mediating Protein Interaction Networks. PLoS Biol. 2005, 3 (12): e405-
Lobley A, Swindells MB, Orengo CA, Jones DT: Inferring function using patterns of native disorder in proteins. PLoS Comput Biol. 2007, 3 (8): e162-
Oldfield CJ, Ulrich EL, Cheng Y, Dunker AK, Markley JL: Addressing the intrinsic disorder bottleneck in structural proteomics. Proteins. 2005, 59 (3): 444-453.
Ferron F, Rancurel C, Longhi S, Cambillau C, Henrissat B, Canard B: VaZyMolO: a tool to define and classify modularity in viral proteins. J Gen Virol. 2005, 86 (Pt 3): 743-749.
Longhi S, Ferron F, Egloff MP: Protein engineering. Methods Mol Biol. 2007, 363: 59-89.
Uversky VN, Oldfield CJ, Dunker AK: Showing your ID: intrinsic disorder as an ID for recognition, regulation and cell signaling. J Mol Recognit. 2005, 18 (5): 343-384.
Ferron F, Longhi S, Canard B, Karlin D: A practical overview of protein disorder prediction methods. Proteins. 2006, 65 (1): 1-14.
Bourhis J, Canard B, Longhi S: Predicting protein disorder and induced folding: from theoretical principles to practical applications. Curr Protein Pept Sci. 2007, 8: 135-149.
Uversky VN, Radivojac P, Iakoucheva LM, Obradovic Z, Dunker AK: Prediction of intrinsic disorder and its use in functional proteomics. Methods Mol Biol. 2007, 408: 69-92.
Vucetic S, Brown C, Dunker K, Obradovic Z: Flavors of protein disorder. Proteins. 2003, 52: 573-584.
Karlin D, Ferron F, Canard B, Longhi S: Structural disorder and modular organization in Paramyxovirinae N and P. J Gen Virol. 2003, 84 (Pt 12): 3239-3252.
Severson W, Xu X, Kuhn M, Senutovitch N, Thokala M, Ferron F, Longhi S, Canard B, Jonsson CB: Essential amino acids of the hantaan virus N protein in its interaction with RNA. J Virol. 2005, 79 (15): 10032-10039.
Ferron FP: Approches bioinformatiques et structurales des réplicase virales. 2005, Marseille: Aix-Marseille II
Llorente MT, Barreno-Garcia B, Calero M, Camafeita E, Lopez JA, Longhi S, Ferron F, Varela PF, Melero JA: Structural analysis of the human respiratory syncitial virus phosphoprotein: characterization of an α-helical domain involved in oligomerization. J Gen Virol. 2006, 87: 159-169.
Chandonia JM, Karplus M: New methods for accurate prediction of protein secondary structure. Proteins. 1999, 35 (3): 293-306.
Chandonia JM: StrBioLib: a Java library for development of custom computational structural biology applications. Bioinformatics. 2007, 23 (15): 2018-2020.
Callebaut I, Labesse G, Durand P, Poupon A, Canard L, Chomilier J, Henrissat B, Mornon JP: Deciphering protein sequence information through hydrophobic cluster analysis (HCA): current status and perspectives. Cell Mol Life Sci. 1997, 53 (8): 621-645.
Dosztanyi Z, Csizmok V, Tompa P, Simon I: IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content. Bioinformatics. 2005, 21 (16): 3433-3434.
Coeytaux K, Poupon A: Prediction of unfolded segments in a protein sequence based on amino acid composition. Bioinformatics. 2005, 21 (9): 1891-1900.
Yang ZR, Thomson R, McNeil P, Esnouf RM: RONN: the bio-basis function neural network technique applied to the detection of natively disordered regions in proteins. Bioinformatics. 2005, 21 (16): 3369-3376.
Garbuzynskiy SO, Lobanov MY, Galzitskaya OV: To be folded or to be unfolded?. Protein Sci. 2004, 13 (11): 2871-2877.
Linding R, Jensen LJ, Diella F, Bork P, Gibson TJ, Russell RB: Protein disorder prediction: implications for structural proteomics. Structure (Camb). 2003, 11 (11): 1453-1459.
Prilusky J, Felder CE, Zeev-Ben-Mordehai T, Rydberg EH, Man O, Beckmann JS, Silman I, Sussman JL: FoldIndex: a simple tool to predict whether a given protein sequence is intrinsically unfolded. Bioinformatics. 2005, 21 (16): 3435-3438.
Linding R, Russell RB, Neduva V, Gibson TJ: GlobPlot: Exploring protein sequences for globularity and disorder. Nucleic Acids Res. 2003, 31 (13): 3701-3708.
Obradovic Z, Peng K, Vucetic S, Radivojac P, Brown CJ, Dunker AK: Predicting intrinsic disorder from amino acid sequence. Proteins. 2003, 53 (Suppl 6): 566-572.
Obradovic Z, Peng K, Vucetic S, Radivojac P, Dunker AK: Exploiting heterogeneous sequence properties improves prediction of protein disorder. Proteins. 2005, 61 (Suppl 7): 176-182.
Kall L, Krogh A, Sonnhammer EL: A combined transmembrane topology and signal peptide prediction method. J Mol Biol. 2004, 338 (5): 1027-1036.
Radivojac P, Iakoucheva LM, Oldfield CJ, Obradovic Z, Uversky VN, Dunker AK: Intrinsic disorder and functional proteomics. Biophys J. 2007, 92 (5): 1439-1456.
This work was supported by the EU VIZIER http://www.vizier-europe.org program (CT 2004-511960) and by the ANR (ANR-05-MIIM-035-02). We gratefully acknowledge the people who provided precious advice on manuscript submission.
This article has been published as part of BMC Genomics Volume 9 Supplement 2, 2008: IEEE 7th International Conference on Bioinformatics and Bioengineering at Harvard Medical School. The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2164/9?issue=S2
The authors declare that they have no competing interests.
PL has designed, developed and implemented the MeDor program and participated to writing the manuscript and the program description material. BC provided advice and funding. SL had the original idea of developing a disorder metaserver and she wrote the manuscript and the description file.