[BiO BB] Association of a protein with a disease

[Ketil Malde]

Pankaj Khurana <pkhurana08 at gmail.com> writes:

> I have a list of ~3000 proteins which are up/down-regulated in a particular
> disease condition.

> 1) what in-silico analysis can be done to identify a biomarker out of them

Do you want to detect the presence of specific proteins?  Way outside my
area of expertise, but surely this is done routinely in medicine?

> 2) Can a prediction model be developed based on them ? So for example to
> check whether a new protein is associated with the disease, just use these
> 3000 and predict likelihood of association with the disease for the  new
> protein.

I would look up the proteins in KEGG, and check out related proteins in
the relevant pathways.

-k
-- 
If I haven't seen further, it is by standing in the footprints of giants