Talk:Harvard:Biophysics 101/2007/Project

Data mining
--Zsun 23:32, 21 March 2007 (EDT)
 * 1) The python file can be found under Bio:Blast:NCBIWWW ...
 * 2) *Documentation of functions here
 * 3) There IS a BLAST API: http://www.ncbi.nlm.nih.gov/BLAST/Doc/urlapi.pdf for directions on how to use
 * 4) *This is CGI based, and judging from the python code may be difficult to hook to

Questions
I'm going to take a look at NCBIWWW.py and see if it's modifiable to give us what we want from BLAST SNP. I'm not sure, because it seems that they're working with an HTML output, rather than XML. Can someone please advise? --TChan 13:00, 20 March 2007 (EST)

Beyond OMIM
I realize that our project is pretty much entirely designed to interface with OMIM and interact with this database, and I think that's a good place to start. Not all of our data-samples will be inherited in a Mendelian manner, thus setting a limit to OMIM's power. This is a concern represented by all of our unassigned tasks. We've discussed the significance of a 'metabolome' and, accordingly, environmental factors on the emergence of specific phenotypes. As we resolve the unassigned tasks, we should facilitate this concern in searching for existing reference databases, creating a new one, or creating a parallel search process given a different input. --CSN 19:35, 20 March 2007 (EST)

I have posted a follow-up on the in-class conversation between me and Shawn, about queries if it is of any interest. --Deniz