genetics-statistics / GEMMA

Genome-wide Efficient Mixed Model Association
https://github.com/genetics-statistics/GEMMA
GNU General Public License v3.0
333 stars 125 forks source link

Difficulties in understanding mvLMM output #236

Closed shafiqnoa closed 4 years ago

shafiqnoa commented 4 years ago

I ran mvLMM for two traits (x and y) adjusting for four additional traits (including the intercept). However, I can not distinguish which one is the estimate for genetic correlation between x and y.

here is the log:

GEMMA Version = 0.98.1 (2018-12-10)

Build profile =

GCC version = 8.2.0

GSL Version = 2.5

Eigen Version = 3.3.5

OpenBlas = OpenBLAS 0.3.2 - DYNAMIC_ARCH NO_AFFINITY Haswell MAX_THREADS=6

arch = Haswell

threads = 1

parallel type = threaded

Command Line Input = /trinity/home/rahmanms/softwares/gemma_0.98.1_linux_static -g /trinity/home/DTR_Shared/DTR_genotypes/HRC_IMP_2017/HRC_r1.1_2MZ/GEMMA_2MZ/HRC_r1.1_TUK_chr22_2MZ_GEMMA.gen.gz -a /trinity/home/DTR_Shared/DTR_genotypes/HRC_IMP_2017/HRC_r1.1_2MZ/GEMMA_2MZ/1000G_P3V5_TUK_chr22_2MZ_GEMMA.annot -k /trinity/home/DTR_Shared/DTR_genotypes/1000G_IMP_2017/1000G_P3V5_2MZ/GEMMA_2MZ/T123_1000g_V8_2MZ_kinship_0941.sXX.txt -c /trinity/home/rahmanms/CWP/IgG_CWP/GEMMA/INPUT_FILES/gemma_covar.txt -lmm 4 -n 9 77 -p /trinity/home/rahmanms/CWP/IgG_CWP/GEMMA/INPUT_FILES/gemma_igg_cwp_all.txt -maf 0.01 -miss 0 -o cwp_igg10_chr22_gwas

Date = Thu Nov 12 15:02:42 2020

Summary Statistics:

number of total individuals = 6921

number of analyzed individuals = 3203

number of covariates = 4

number of phenotypes = 2

number of total SNPs/var = 524544

number of analyzed SNPs/var = 106465

REMLE log-likelihood in the null model = -6081.36

MLE log-likelihood in the null model = -6084.94

REMLE estimate for Vg in the null model:

0.69833 0.0375916 0.065652

se(Vg):

0.0321777 0.0108328 0.0072974

REMLE estimate for Ve in the null model:

0.273633 -0.00751629 0.120799

se(Ve):

0.0184782 0.00813352 0.00677273

MLE estimate for Vg in the null model:

0.69775 0.0375614 0.0375614 0.0655758

se(Vg):

0.0321219 0.0108154 0.00728674

MLE estimate for Ve in the null model:

0.273107 -0.00751642 -0.00751642 0.120644

se(Ve):

0.018433 0.008118 0.0067622

estimate for B (d by c) in the null model (columns correspond to the covariates provided in the file):

0.0940317 -0.0878004 0.166091 -0.0150725 0.406611 -0.120677 0.0680489 -0.0198323

se(B):

0.0752651 0.0684905 0.0199194 0.0139538 0.0320747 0.0288643 0.00833621 0.00586676

Computation Time:

total computation time = 24.7901 min

computation time break down:

time on eigen-decomposition = 0.133633 min

time on calculating UtX = 1.10756 min

time on optimization = 18.718 min

xiangzhou commented 4 years ago

Vg is the genetic covariance. You can obtain genetic correlation based on this covariance estimate.