### Post by rodrigomarin on Jun 21, 2022 17:07:23 GMT

Hi there, here's my code.

Reading IDs of the GRM from [GRM.grm.id].

5839 IDs read from [GRM.grm.id].

Reading the GRM from [GRM.grm.bin].

GRM for 5839 individuals are included from [GRM.grm.bin].

Reading phenotypes from [ALL_BS_PHENO.txt].

Non-missing phenotypes of 5959 individuals are included from [ALL_BS_PHENO.txt].

Reading quantitative covariates from [ALL_QCOVAR.txt].

1 quantitative covariate(s) of 3271 individuals read from [ALL_QCOVAR.txt].

Reading discrete covariate(s) from [ALL_COVAR.txt].

2 discrete covariate(s) of 5958 individuals are included from [ALL_COVAR.txt].

Assuming a disease phenotype for a case-control study: 1464 cases and 1688 controls

Note: you can specify the disease prevalence by the option --prevalence so that GCTA can transform the variance explained to the underlying liability scale.

1 quantitative variable(s) included as covariate(s).

2 discrete variable(s) included as covariate(s).

3152 individuals are in common in these files.

Performing REML analysis ... (Note: may take hours depending on sample size).

3152 observations, 4 fixed effect(s), and 2 variance component(s)(including residual variance).

Calculating prior values of variance components by EM-REML ...

Updated prior values: 6.88342 1.39152

logL: 2689.64

Running AI-REML algorithm ...

Iter. logL V(G) V(e)

1 -2472.78 -6.94324 -6.34323

2 -3621.88 10.95667 -290.75664

3 -8857.28 18.72038 -520.38176

4 -9777.34 2.50090 -169.48828

5 -8103.76 2.48988 -164.10123

6 -8018.68 2.52059 -173.16718

7 -8072.39 3.02644 -210.13694

8 -8395.09 41.74918 1320.64335

9 -11352.29 76705.98330 -5133572.46655

Error:

the X^t * V^-1 * X matrix is not invertible. Please check the covariate(s) and/or the environmental factor(s).

There are 4 studies and only 2 have a qcovariate, almost half of the dataset has no qcovar and those were written as "NA".

Beside a likely colineality problem among covariates, why does the software cutting of animals with only qcovariates and doesn't count them as NAs ?

Reading IDs of the GRM from [GRM.grm.id].

5839 IDs read from [GRM.grm.id].

Reading the GRM from [GRM.grm.bin].

GRM for 5839 individuals are included from [GRM.grm.bin].

Reading phenotypes from [ALL_BS_PHENO.txt].

Non-missing phenotypes of 5959 individuals are included from [ALL_BS_PHENO.txt].

Reading quantitative covariates from [ALL_QCOVAR.txt].

1 quantitative covariate(s) of 3271 individuals read from [ALL_QCOVAR.txt].

Reading discrete covariate(s) from [ALL_COVAR.txt].

2 discrete covariate(s) of 5958 individuals are included from [ALL_COVAR.txt].

Assuming a disease phenotype for a case-control study: 1464 cases and 1688 controls

Note: you can specify the disease prevalence by the option --prevalence so that GCTA can transform the variance explained to the underlying liability scale.

1 quantitative variable(s) included as covariate(s).

2 discrete variable(s) included as covariate(s).

3152 individuals are in common in these files.

Performing REML analysis ... (Note: may take hours depending on sample size).

3152 observations, 4 fixed effect(s), and 2 variance component(s)(including residual variance).

Calculating prior values of variance components by EM-REML ...

Updated prior values: 6.88342 1.39152

logL: 2689.64

Running AI-REML algorithm ...

Iter. logL V(G) V(e)

1 -2472.78 -6.94324 -6.34323

2 -3621.88 10.95667 -290.75664

3 -8857.28 18.72038 -520.38176

4 -9777.34 2.50090 -169.48828

5 -8103.76 2.48988 -164.10123

6 -8018.68 2.52059 -173.16718

7 -8072.39 3.02644 -210.13694

8 -8395.09 41.74918 1320.64335

9 -11352.29 76705.98330 -5133572.46655

Error:

the X^t * V^-1 * X matrix is not invertible. Please check the covariate(s) and/or the environmental factor(s).

There are 4 studies and only 2 have a qcovariate, almost half of the dataset has no qcovar and those were written as "NA".

Beside a likely colineality problem among covariates, why does the software cutting of animals with only qcovariates and doesn't count them as NAs ?