Expression Prediction

[jjiang92292]

I utilize the FUSION.compute_weights.R and FUSION.profile_wgt.R to calculate specific weight for my gene list and to generate the best regression models for each genes with input genotype of ±100KB around genes. Next I would like to utilize the best models to predict expression level for each genes also by the ±100KB variants, which helps me to reveal the cis-expression component of each gene. In other word, I want to use FUSION to predict genes expression instead of conducting summarized TWAS analysis. How could I reach this?