I have been trying to filter the subnetwork associated with specific cells example CD14 from homo sapiens PPI network.
How to filter from the homo sapiens PPI network?
So that the resulting network is more specific for CD14 cells.
And how to perform unbiased validation of the CD14 network.
A way to do this is to use expression or proteomics data from the cell line and remove from the network all genes not expressed in that cell line. I am not sure what you mean by unbiased validation but in most cases, independent validation means doing actual experiments to confirm computational findings. I think it is currently difficult to try to confirm that the whole network you predict is specific of a particular cell line because many interactions will not be specific even though they occur in the cell line of interest and interactions being dynamic, the network you predict is not going to actually exist but it'll be, as its source data, an aggregation of various cell states. However, you could still experimentally test whether some interactions from your network do occur in the cell line.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Thank you for your suggestion. If I understood well the below flowchart is what you meant. Example GEO -> CD14 (human) ->Expression (How to fish out genes which are not expressed)
Kindly correct me if I am wrong.
Yes. The trick is simply to remove from the network proteins that are not expressed in the cell line. You could also use the ENCODE cell line data to figure out genes that are only/mostly expressed in CD14 cells, see e.g. ENCODE data in the Expression Atlas.
I need to search inside encode for the cell type of my interest and do the needful... Thank you once again.