Open rbloom5 opened 9 years ago
For example - others have shown that VH4 is used more often in MS patients than in controls. It would be interesting to see if that is true in our data, and find other things like that.
I have some basic V, D, and J segment usage analysis up and running (currently on the feature-analysis branch). My next step will be to tackle CDR length and composition stuff. I need to find some good examples in the literature. I'm definitely open to any suggestions for where to take this.
Yea, we are definitely venturing into uncharted territory a bit here. I think the following references are pretty good - but we can definitely talk more about it this afternoon
http://www.ncbi.nlm.nih.gov/pubmed/23390249 http://www.ncbi.nlm.nih.gov/pubmed/25100740 http://www.ncbi.nlm.nih.gov/pubmed/23160197
as well as a lot of this stuff: http://www.hackbio.com/cv/
Clustering, frequency statistics etc. on simple features like V segment usage, J segment usage, CDR length, composition etc etc. The goal would be to try and find simple features that distinguish MS from healthy patients.