How to add more fixed effects to the MLMA model?

Mema
New Member

Posts: 6

How to add more fixed effects to the MLMA model? Oct 27, 2015 9:27:37 GMT

Quote

Post by Mema on Oct 27, 2015 9:27:37 GMT

Hello,

I am trying to run this model using GCTA: gcta64 --mlma

Y (disease binary trait) = Xb (fixed effects – calving year and calving month) + Ma (SNP 50k snp-chip) + Zg (polygenic effect fitted through pedigree or GRM-genomic relationship matrix) + e

And after reading the manual for GCTA cnsgenomics.com/software/gcta/mlmassoc.html , I still have 2 questions:

1- How to include the 2 fixed effects (which are calving year and calving moth) in the running-code and in the input file?
2- How to replace the GRM by the Additive relationship matrix (using the Pedigree)?

N.B i am using these example gcta64:
gcta64 --mlma --bfile gwas_bin_data --grm gwas_bin_data pheno binaryPheno.txt out gwas_bin_data --thread-num10

Thanks in Advance

Best regards
Medhat

Last Edit: Oct 27, 2015 9:39:31 GMT by Mema

Jian Yang
Administrator

Posts: 362

How to add more fixed effects to the MLMA model? Oct 27, 2015 22:52:40 GMT

Quote

Post by Jian Yang on Oct 27, 2015 22:52:40 GMT

Re 1) Using --covar or --qcovar to include fixed covariates.

Re 2) Making the pedigree relatedness matrix, saving it in *.grm.gz format (see gcta.freeforums.net/thread/175/gcta-estimating-genetic-relationship-using for the format of a *.grm.gz file), reading the pedigree GRM using the option --grm-gz.

Last Edit: Oct 27, 2015 22:53:16 GMT by Jian Yang

domi
New Member

Posts: 3

How to add more fixed effects to the MLMA model? Sept 7, 2021 16:05:18 GMT

Quote

Post by domi on Sept 7, 2021 16:05:18 GMT

By adding calving year and calving month as fixed covariad, you should not use a header. But how does the program know that the numbers in these colomns present the calving year and calving month?
I am just wondering because i did the same putting my candidate SNP (Genotype coded as C/C = 1, C/G = 2 and G/G = 3) and sex into my covar-file and age, weight and high in my qcovar-file and it worked perfectly