thank you so much. so people usually use a prediction db rather than a primary database from literature? I am validating my thoughts with what people do, I am a newbie :)
Hi everyone
Do you know what is the best PPI network data base for human? I have some genes that we doubt they are related to tumor, but we need to do some more investigations? I use Biogrid, wonder if everyone uses anything else?
Thanks
3 answers
How about STRING: functional protein association networks, which appears to be quite popular recently.
It allows you to work with your own data.
Well, one can build what's called a co-expression network, which is essentially just representing correlations between your genes or proteins. In this way, for example, one can remove edges hat fall below a certain correlation threshold (like absolute Pearson r < 0.8). In this regard, I posted a tutorial here: Network plot from expression data in R using igraph
One can also use information from online databases and published literature, which is STRING's default mode of functioning.
For physical protein-protein interactions, I recommend working with aggregated data such as iRef index (or compiling such an aggregated data set yourself if you need something more up-to-date) as this is more exhaustive. For functional association, you can combine different types of data as for example done by GeneMANIA. For an introduction on how this works, have a look at my paper here.
I think these articles can be very useful for you to begin to better understand how to build your networks:
- Interactome networks and human disease.
- Using graph theory to analyze biological networks
- Construction and analysis of protein–protein interaction networks
- Network based meta-analysis prediction of microenvironmental relays involved in stemness of human embryonic stem cells
Best regards,
Leite
Log in to answer this question.
