GSoC 2016 – Week 1

First Week 🙂

Here it goes!!

After discussion with my mentor, we decided to start working on GPML format which is extensively used as a file format for biological pathways.

We decided to work with sample files available on http://www.wikipathways.org/index.php/WikiPathways. To parse the file we used https://docs.python.org/2/library/xml.etree.elementtree.html and extracted all the information available like various data nodes, interactions, groups, etc.

One major difference is that in formats like cytoscape JSON the edges(interaction) are defined between two nodes but in the case of GPML they are defined only by the coordinates. So we need to create an initial layout for the network so it can be rendered properly by GraphSpace.

GraphSpace uses cytoscape web in the backend to render graphs. To solve the problem of creating groups we can use http://cytoscapeweb.cytoscape.org/tutorial#section/Compound_Nodes provided by cytoscape web.

By next week I plan to have a working prototype of the GPML upload function.

Cheers!

Advertisements

GSoC 2016 – Week 0/ Community bonding period

The official coding period of GSoC 2016 starts tomorrow, 23rd May. Yay!!

I’ll start working on creating a parser for GraphML format. Users of GraphSpace will be able to upload their graphs in GraphML format without having to convert it to JSON or cyjs (cytoscape JSON).

GraphML is based on XML, hence we’ll use the standard XML python parser https://docs.python.org/2/library/xml.etree.elementtree.html

Implementing this may require changing the GraphSpace JSON spec, which will be further discussed with my mentors.

We also make sure to extensively document the new functionality (using the PyDoc format).

My mentors Murali, Ted and Divit will be able to view a dev version of GraphSpace on the dev server shared with them.

I’ll regularly update this blog with the progress through the summer. 🙂

Cheers!