d-Omix: A Mixer of Generic Protein Domain Analysis Tools

Data

Data tab allows a user to upload raw InterProScan resulted files as input for protein domain analysis tools including Tree, Graph, Alignment, and Interaction.

The click on [Example] will perform an automatic loading of exampled data sets.

With at least one uploaded data set, tabs for domain analysis tools will be activated. For each uploaded data set, the user needs to specify a name for it. The

user may merge data sets together for the comparative analyzes. In this example, we merged Gibberellin and Autonomous data sets together.

Tree

When Tree tab selected, the following page will be shown. Data file, Domain search tools, and Operation sections will be the same for all analysis tools.

- Data file section will contain all data files selected and/or merged from the Data tab. These data files could be updated by going back to customize

the data files in Data tab. Note that the current results will be discarded with the new selected and/or merged data sets.

- Domain search tools section lists all domain search tools that the user could select for the analysis. Note that some of the selected search tools

might not produce results, as there is no input data for those search tools.

- Operation section simply contains button to start the analysis. The reset button allow the user to redo the analysis with changed data files and/or

domain search tools and/or specialized parameters as in case of Tree and Alignment.

Tree option section is specialized for Tree analysis. It allows the user to compare domain architecture (DA)-based trees with CLUSTALW-based trees.

The following figure shows the results after Tree building.

Tabs in result section correspond to selected files in Data file section. In this example shows DA-based tree built from the merged data file between Gibberellin and Autonomous data sets. The click on “View domainAC” button will redraw the DAs of proteins using domain accession numbers (AC) instead of domain identifiers (ID) as shown in the following figure.

The click on icon will send the current tree to PhyloWidget [1] for further viewing and editing the resulted tree.

Graph

Domain graph mainly computes the versatility and abundance [2] of protein domains. When Graph tab selected, the same three sections: Data file, Domain search tools, and Operation will be shown. After selecting some data files and domain search tools, the click on “Build graph” button will generate domain graphs. The following figure shows domain graph with direction built from the merged data file between Gibberellin and Autonomous data sets.

The click on “View domainAC” button on the left will redraw domain graphs using domain accession numbers (AC) instead of domain identifiers (ID) as shown in the following figure.

If the domain graph is large, the user may click on icon to send the current graph to ZGRViewer [3] for easy navigation and smooth zoomable features.

Alignment

The Alignment tool allows the user to compare the similarities between proteins based on the similarities of their protein domain architectures (DAs). In this page, besides the same three Data file, Domain search tools, and Operation sections, the user also needs to specify if a selected data set is a source or target which is respectively analogous to the queried sequences and searched against database in BLAST. Two additional specialized parameters: DA score and Hit limit enable the user to limit the number of similar proteins in the alignment result. The lower DA score the more similar DAs. Two proteins have exactly the same DAs if DA score between them is zero. The following figure shows an example of Alignment result.

Interaction

Interaction tool builds a tentative protein network directly derived from domain-domain interactions from DOMINE [4]. After selecting some data files and domain search tools, the click on “Build tentative protein network” button will generate the results as shown in the following figure. The shown tentative protein network is built from the merged data file between Gibberellin and Autonomous data sets. An edge between two proteins represents the availability of known domain-domain interactions forming the DAs of the two proteins.

The user may filter the protein network according to the confidence levels: iPfam ME 3did, HC, and LC of domain-domain interactions from DOMINE. The click on an edge between two protein nodes will refresh the DA alignment detail on the right with the DAs of these proteins and a click on a protein domain with domain-domain interaction information will highlight the domain and its interacting partners as shown in below figure. Similar to domain graph, if the tentative protein network is large, the user may click on icon to send the current network to ZGRViewer[3] for easy navigation and smooth zoomable features.

Prepare InterProScan file

A user with only protein sequences could still utilize d-Omix web server with an additional step for raw InterProScan preparation. This option is located in Data tab. The click on [Prepare InterProScan file] will show the following page. The user may cut and paste or upload their protein sequences. In this figure, protein sequences in FASTA format comes from the click on [Example] in this section. The user may select specific domain search tools for InterProScan [5] running.

The figure below shows the result page of InterProScan file preparation. The click on “Add to d-Omix” button will automatically load the resulted file as an input file for d-Omix with a default but customizable name.

REFERENCES:

1. Jordan GE, Piel WH (2008) PhyloWidget: web-based visualizations for the tree of life. Bioinformatics: btn235.

2. Vogel C, Teichmann SA, Pereira-Leal J (2005) The Relationship Between Domain Duplication and Recombination. Journal of Molecular Biology 346: 355.

3. Pietriga E (2005) A Toolkit for Addressing HCI Issues in Visual Language Environments. IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC) 00: 145-152.

4. Raghavachari B, Tasneem A, Przytycka TM, Jothi R (2008) DOMINE: a database of protein domain interactions. pp. D656-661.

5. Quevillon E, Silventoinen V, Pillai S, Harte N, Mulder N, et al. (2005) InterProScan: protein domains identifier. Nucl Acids Res %R 101093/nar/gki442 33: W116-120.