Entering edit mode
6.8 years ago
amroumouna
•
0
I used R STRINGdb package to get neighbors of oncogenes and suppresor genes for Cancer . I have generated 5826 mappiung identifier , and 3916 interaction . but I need a large network I need to build a network of 200k proteins to use it in Neo4j. Is there anyway to do so , to build a network of 200k out of the mappings file
200k proteins ? It's unlikely you'll find that many human proteins in the human genome. The human genome has ~20k protein-coding genes, 200k proteins mean an average of 10 proteins per gene. Even if that were true, they wouldn't be all in the public databases given that ~20% of human protein coding genes are completely uncharacterized.
what is the largest network I can get for human genome them ? approximatively
Please use the 'add reply' button to reply to a comment. This keeps the discussion organized and avoids marking the question as answered when it is not.
If you want a comprehensive network of human protein-protein interactions, you could use iRefIndex (you'd have to filter out the non-human interactions). If you want networks representing other types of relations then you'd have to define which ones you're interested in. Combining protein-protein interactions in human, protein interactions transferred by orthology from model organisms, relations based on biological process GO terms annotations and text mining à la iHOP, you would get a network covering ~14000 protein-coding genes.