Hi Tiff,
Thanks for your interest. Idea 4 is essentially doing ID
conversion. Taking user's input ID list, do the conversion by
calling a web service (via
mygene.info), and then display the
conversion result back to the user. There is no need to store the
result in BioGPS, no database needs to be involved. Here is an
example:
user's input:
1053_at
117_at
121_at
and this user tells us these are reporter IDs, and wants matching
GeneSymbol, GeneID, GeneName back. By calling
mygene.info web
service, we can get all these needed data back, then we then display
the conversion result like this in user's browser:
reporter symbol geneid name
1053_at RFC2 5982 replication factor C (activator
1) 2, 40kDa
117_at HSPA6 3310 heat shock 70kDa protein 6 (HSP70B')
121_at PAX8 7849 paired box 8
The second feature is doing the similar ID conversion as well. The
difference is that, instead of users pasting in a ID list, users
upload a multi-column table file, and specify which column is the ID
column for the ID conversion. Then the same conversion will be done
based on this column.
This project will implement a standalone web application.
It requires both server-side backend and browser-side user
interface. Yes, we require students to implement the backend in
Python. And you may consider to use a web framework like Tornado,
Django, etc. Probably even more important is to build a smooth user
interface, which requires the knowledge of Javascript/HTML/CSS. If
you are a Python beginner, it will be a good opportunity for you to
gain experience on building web application in Python. If you have
experience with Javascript/HTML/CSS before, I think you will be a
good-fit.
For the application form, you may use this one: "Advanced
gene annotation tools for BioGPS". In your application, you should
describe your implementation plan for these two features. It's not
required to come up new ideas in your application, but new ideas are
always welcome.
Feel free to ask if any further questions.
Chunlei