Set up the data

# Load required packages
library(mashr); library(ExtremeDeconvolution); library(flashr2)

Loading required package: ashr

# read data
data = readRDS('../data/ImmuneQTLSummary.4MASH.rds')
data$max$se = data$max$beta/data$max$z
data$null$se = data$null$beta / data$null$z

# set parameters
K = 10
P = 5
vhat = 1

Fit mash model

Using a random null set, the cell below computes the weights for input covariance matrices (Covariance matrices plus loading) in MASH mixture. The output contains matrix of log-likelihoods as well as weights learned from the hierarchical model.

if (vhat == 1) {
  V = cor(data$null$z[which(apply(abs(data$null$z),1, max) < 2),])
} else {
  V = diag(ncol(data$null$z))
}

mash_data = mashr::set_mash_data(Bhat = as.matrix(data$null$beta), 
                                 Shat = as.matrix(data$null$se), 
                                 V=as.matrix(V), alpha=1)

Ulist = readRDS('../output/ImmuneEZ.U.center.xtx.Load.K10.P5.rds')$Ulist

saveRDS(mashr::mash(mash_data, Ulist, outputlevel = 1), 
        paste0('../output/ImmuneEZ.V',vhat,'.center.mash_model.Load.K',K,'.P',P,'.rds'))

 - Computing 193301 x 1008 likelihood matrix.
 - Likelihood calculations took 26.05 seconds.
 - Fitting model with 1008 mixture components.

Warning in REBayes::KWDual(A, rep(1, k), normalize(w), control = control): estimated mixing distribution has some negative values:
               consider reducing rtol

Warning in mixIP(matrix_lik = structure(c(0.256254290186069,
0.750668812603541, : Optimization step yields mixture weights that are
either too small, or negative; weights have been corrected and renormalized
after the optimization.

 - Model fitting took 1142.36 seconds.

Posterior inference

Applying hyperparameters learned from the training (the null) set to the test (the top eQTL) set, the cell below computes posterior quantities.

U.m = readRDS('../output/ImmuneEZ.V1.center.mash_model.Load.K10.P5.rds')

if (vhat == 1) {
  V = cor(data$null$z[which(apply(abs(data$null$z),1, max) < 2),])
} else {
  V = diag(ncol(data$null$z))
}

mash_data = mashr::set_mash_data(Bhat = as.matrix(data$max$beta), 
                                 Shat = as.matrix(data$max$se), 
                                 V = as.matrix(V), 
                                 alpha = 1)

saveRDS(mash_compute_posterior_matrices(U.m, mash_data), 
        paste0('../output/ImmuneEZ.V',vhat,'.center.mash_posterior.Load.K',K,'.P',P,'.rds'))

FIXME: 'compute_posterior_matrices' in Rcpp does not transfer EZ to EE

Session information

sessionInfo()

R version 3.4.3 (2017-11-30)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS High Sierra 10.13.2

Matrix products: default
BLAS: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] flashr2_0.4-0            ExtremeDeconvolution_1.3
[3] mashr_0.2-4              ashr_2.1-27             

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.14      knitr_1.17        magrittr_1.5     
 [4] REBayes_1.2       MASS_7.3-47       doParallel_1.0.11
 [7] pscl_1.5.2        SQUAREM_2017.10-1 lattice_0.20-35  
[10] foreach_1.4.4     plyr_1.8.4        stringr_1.2.0    
[13] tools_3.4.3       parallel_3.4.3    grid_3.4.3       
[16] rmeta_2.16        git2r_0.20.0      htmltools_0.3.6  
[19] iterators_1.0.9   assertthat_0.2.0  yaml_2.1.16      
[22] rprojroot_1.2     digest_0.6.13     Matrix_1.2-12    
[25] codetools_0.2-15  evaluate_0.10.1   rmarkdown_1.8    
[28] stringi_1.1.6     compiler_3.4.3    Rmosek_8.0.69    
[31] backports_1.1.2   mvtnorm_1.0-6     truncnorm_1.0-7