Multivariate GEMMA hangs forever #243
Description
On a simulated dataset with 1000 samples, 10000 SNPs, and 2 traits, multivariate GEMMA ran for over an hour. Is this reasonable?
My commands are:
# generate relatedness matrix
gemma -bfile multivariate_2traits -gk 1 -o multivariate
# run gemma
gemma -bfile multivariate_2traits -k output/multivariate.cXX.txt -maf 0.0000001 -lmm 4 -n 1 2 -o gemma.polygenic.result
Output: (the 100% shows up almost immediately, but hangs forever and never displays the "done").
GEMMA 0.98.4 (2021-01-29) by Xiang Zhou and team (C) 2012-2021
Reading Files ...
## number of total individuals = 1000
## number of analyzed individuals = 1000
## number of covariates = 1
## number of phenotypes = 2
## number of total SNPs/var = 10000
## number of analyzed SNPs = 10000
Start Eigen-Decomposition...
REMLE estimate for Vg in the null model:
23.5963
6.2719 23.8002
se(Vg):
4.4394
2.7440 5.4686
REMLE estimate for Ve in the null model:
2.6355
0.0002 0.0002
se(Ve):
1.3386
0.8088 1.6734
REMLE likelihood = -4988.2498
MLE estimate for Vg in the null model:
25.3766
6.2899 23.8008
se(Vg):
5.8054
0.8981 1.0649
MLE estimate for Ve in the null model:
2.0757
0.0011 0.0000
se(Ve):
1.7524
0.0009 0.0000
MLE likelihood = -4981.3187
================================================== 100%
If I analyze the traits separately, the analysis finishes in ~5 seconds. I assume this is a bug but I'm not sure.
For reproducibility, my test PLINK files are zipped and attached.