The project result becomes a part of the paper
Carlin, D.E., Fong, S.H., Qin, Y., Jia, T., Huang, J.K., Bao, B., Zhang, C. and Ideker, T., 2019. A fast and flexible framework for network-assisted genomic association. Iscience, 16, pp.155-161.
Slides Link
https://docs.google.com/presentation/d/1to310NDKKWqgWbuMVb4zgLRPSoP3ZXkGajnbKoFuB2I/edit?usp=sharing
Overleaf https://www.overleaf.com/read/prjnjmhbwxyp https://www.overleaf.com/16700835nghxcnnkxwtm#/64031895/
SNPs' statistic scores are obtained.
NN is built with following structure:
SNPs -> Gene -> Weighted Gene -> SNPs
SNPs are sampled for NN trainning and validation
The Weighted Gene layer is taken out and the weight is used for gene ranking. Done.
Use the top-hitted SNPs pvalue as the gene value.
Run the GWAB or NetWAS to do the network propagation.