GSoC/GCI Archive
Google Summer of Code 2012

Crowdsourcing Biology

Web Page:

Mailing List:!forum/crowdbio

Our team develops new technologies and tools to enable large communities of scientists to collaboratively address biological challenges of massive scale.  We are part of the the Scripps Research Institute in La Jolla, California, one of the largest non-profit research organizations devoted to biomedical research.  We operate at the exciting intersection between biological and computational research.


All of our projects revolve around the theme of "crowdsourcing" -- directly engaging scientists to collaboratively achieve projects of massive scale.  We currently have three project areas, each of which have plenty of room for GSoC contributors:

  • Gene Wiki: comprehensively organize knowledge of all human genes
  • BioGPS: provide a gene annotation portal for aggregating gene-centric online content
  • Biological games: build games that build scientific knowledge through game play

Students, please see our ideas page and visit our Google Group for more details!


NOTE: If you'd like to apply to be a mentor, please email us at first with more details.  Requests from people who haven't discussed it with us beforehand will be rejected.


  • A Scalable and Efficient Storage System of GeneLists and Enrichment Score Computation I plan to create a fully operational REST interface which should map directly to CRUD operations which act on a database of gene lists. The RESTful application will be written in python, and interact directly with a MongoDB store which contains a collection of all gene lists. A computation engine will be created to calculate statistical enrichment using an all to all comparison of gene lists. By accomplishing this project, users will have an alternate method to discovering similar genes to ones they are interested in.
  • BioGPS:JQuery-based BioGPS gene-report layout canvas My main aim is improving the Front end for improved User Interaction. What the project demands is an "Interactive jQuery based canvas" part of the site, people have more places to contribute, and engage people when they are benefiting from the site. The following application is an abstract of what I want to implement and hence propose the same.
  • Data Visualization of Gene-Disease Links Using Graphical Networks I intend to visualize gene-disease links by generating graphical networks by using the Cytoscape Web( on the web platform. I intend to use Flash because of the low processing-high performance quality of Adobe Flash. Cytoscape Web uses javascript to dynamically manipulate the Flash object thus, providing both the speed and performance of Flash as well as the dynamic nature of data manipulation using javascript. The interface will be designed in a way to avoid cumbersome navigation through a huge graph of data. There will be options to set the size of the chart as well as zoom and search options to navigate around the graph with ease. The data for the visualization will be obtained using the powerful semantic query search of Gene Wiki+. I have already gone through the API at For eg, contain XML data regarding Cancer. This XML data will be visualized using graph networks. Features I intend the interface to have at the end of my project 1) Movable nodes representing the specific disease and the associated genes, which can be positioned according to the biologist's needs. Clicking of these nodes will retrieve information regarding the term and will be displayed alongside the interface using Ajax and the semantic query search. 2) Color coded connections representing various relationships. 3) Exporting of the graphical network at any point as an image for further use by the biologist. 4) Options to add various nodes and connections based on the biologist's needs in which case further data regarding the relationship will be obtained and displayed using Ajax. 5) Options to delete nodes and connections and to vary size of the panel as seen fit by the biologist. 6) Suggestions regarding relationships and addition of nodes to the biologist by dynamically analyzing the nodes on the interface. 7) To negate the need of a textbox based search as currently present at
  • Web annotation interface My name is Maximilian Ludvigsson. This project is a web interface for letting users annotate information on the various plugins in BioGPS.