bayesR
Bayesian hierarchical model for complex trait analysis
(https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1004969)
Update 01/04/2021
Various new features and improvements:
- further reduced memory requirements
- inclusion of covariates
- grouped effects models to fit more complex models (e.g. partioning of variance)
- flat input files (not supported in bayesRv2)
- fitted values
- prediction of phenotypes
- unified source code
Previous release moved to folder /old
Quick start
Clone:
git clone https://github.com/syntheke/bayesR.git
Compile:
in the src folder
gfortran –o bayesR –O2 -cpp RandomDistributions.f90 baymods.f90 bayesR.f90
gfortran –o bayesRv2 –O2 -cpp –Dblock –fopenmp RandomDistributions.f90 baymods.f90 bayesR.f90
ifort –o bayesR –O3 -fpp RandomDistributions.f90 baymods.f90 bayesR.f90
ifort –o bayesRv2 –O3 -fpp –Dblock –openmp –static RandomDistributions.f90 baymods.f90 bayesR.f90
Run:
bayesR -bfile simdata -out simout
Example1
Example from the 14th QTL-MAS workshop.
bayesR -bfile example/simdata -out simout -numit 10000 -burnin 5000 -seed 333
Example2
Genome position specific priors
bayesR –bfile simdata2 –out simout2 –numit 10000 –burnin 5000 –seed 333 -n 2 -snpmodel mod2 -segment seg
Example3
Grouped effects with mixture priors
bayesR –bfile simdata2 –out simout3 –numit 10000 –burnin 5000 –seed 333 -n 2 -snpmodel mod3 -segments seg -varcomp var3
Help:
bayesR –help