GSoC Projects

Google Summer Of Code sigVIZ Project Description

Structural Systems Biology: Modeling and Visualizing Cell Signaling Networks to Improve Communication and Hypothesis Generation with sigViz

This project aims to develop software, sigViz, that researchers can use to create interactive dynamic representations of cell signaling pathways that contain accurate structural details for improved communication, hypothesis generation, and systems simulation. 

The sigViz app will function in a variety of popular interfaces, including web browsers, to enable researchers from any background to easily access and use the tools with no special training. A graphic user interface designed for quantitative biologists will produce outputs that target diverse audiences and purposes (e.g., journals, slides, websites, and thought/virtual experiment tools).

The GSoC sigViz project programmer projects are designed to pare down, clean up, merge and build upon generalized molecular viewer and bioinformatics visualization platforms (including CytoscapeuPyePMV, and Chimera) currently developed at UCSF and Gladstone to create sigViz as an intuitive cell signaling visualization and modeling app.

GSoC candidates may select from the following three projects to best match their interests and/or abilities:

  1. Create a metabolic flux tool for the Cytoscape App Store targeted to cell biologists
    1. Description: This tool will add coarse flux modeling based on dose response curve theory to the cell signaling networks available in the Cytoscape database. The tool will function similar to PathExplorer (from Cytoscape App Store), but will have a structure and user interface optimized for quantitative biologists rather than network visualization experts. A second aspect of the project will access community currated annotation and weighting to enable users to filter cell signaling networks to any level of detail.
    2. Required skills: Java, JavaScript, Python
    3. Recommended experience: Cytoscape programming experience
  2. Cell Signaling Pathway Structural Visualization
    1. Description: Molecules, represented as nodes in Cytoscape network diagrams, will become annotated with 3D structural representations from a variety of sources to enable visualization of all of the players in structure-annotated-pathways. Molecular models will be constructed in realtime (likely via the existing structureViz or uPy apps) and rendered into the network diagrams at appropriate resolutions and absolute size to enhance effective and efficient study through web browser interfaces. The 3D structural representations annotations will be generated automatically and manually, and will be curated by the Cytoscape community by plugging into existing Cytoscape wiki-mediating mechanisms.
    2. Required skills: JavaScript, Python
    3. Recommended experience: webGL, uPy, Unity3D, Chimera, HTML5, database, StructureViz
  3. Cell Signaling Pathway Animation/Simulation 
    (contingent upon co-development of project 2 above)
    1. Description: As described in the previous project, molecules, represented as nodes in Cytoscape network diagrams, will become annotated with 3D structural representations from a variety of sources to enable visualization of all of the players in structure-annotated-pathways. In the project, a programmer will design a system that allows the biology community to annotate the edges of cell signaling networks with orders of operation (absolute and probabilistic) and interaction types to visualize animated interactions of the nodes along the pathways. Such animations will include physical binding events as mocked in this flash animation, as well as graphical/metaphorical representations for non-physical interactions prevalent in many cell signaling pathways like genetic interactions.
    2. Required skills: JavaScript, Python
    3. Recommended experience: webGL, uPy, Unity3D, Chimera, HTML5, database, game design, physics simulation e.g., Bullet 3D

Figure 1: Traditional signaling visualization tool outputs compared to planned sigViz outputs

A. A representative diagram demonstrates the disconnect between good computational work and clear visualization and communication. (Saez-Rodriguez, et. al, Cancer Res August 15, 2011 71;5400).

B–C. sigViz interactive models will allow the user to shuffle and modify complex circuits and networks in an intuitive way for use as a thinking tool. Animated flow will clearly show how each node gets activated or inhibited and the overall results of activation or partial activations at each node as in the change from B to C. Structural details and cellular contexts can be added for enhanced use as a thinking and communication tool.

D. sigViz will further allow users to create diagrams and movies for communication and outreach that incorporate the physical realities of protein-protein interactions and best methods of good design. EFG signaling cascade by Johnson for the textbook, Cell Biology 2e, by Tom Pollard and Bill Earnshaw, 2007.

Please view an animated version of a ß-adrenergic signaling mechanism (made in 2001 by biovisualization mentor Johnson) at: http://www.grahamj.com/PollardCellBiology/1eCD/data/animations/Flash2903.html

Subpages (1): fluxViz
Comments