Error with binary GWAS encoding

[hde08]

Hello,

Thank you very much for this great tool. I have been using GEMMA without issues for a while but now wanted to do a GWAS using a binary trait, as I have seen it is supported in GEMMA manual, but this does not seem to work properly.

I have encoded the binary trait as 0/1 (control/cases) as outlined in the manual. However when running the GWAS model (here lmm but the same occurs irrespective of model type), all individuals encoded as 0 are considered as missing and GEMMA crashes (see below)

 ```##GEMMA 0.98.5 (2021-08-25) by Xiang Zhou, Pjotr Prins and team (C) 2012-2021
##Reading Files ...
##number of total individuals = 624
##number of analyzed individuals = 107
##number of covariates = 1
##number of phenotypes = 1
##number of total SNPs/var        =   440918
##number of analyzed SNPs         =   440669
##Start Eigen-Decomposition...
##pve estimate =2.4726e-06
##se(pve) =-nan
**** WARNING: Brent did not converge               5%
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
**** WARNING: Brent did not converge
``` 

107 corresponds to the number of cases only, the total sample size is 211. I managed to run it by encoding the binary trait as 1/2 (control/case) but I suspect GEMMA is not running the model correctly (see below)

 ```##
## GEMMA Version    = 0.98.5 (2021-08-25)
## Build profile    = /gnu/store/8rvid272yb53bgascf5c468z0jhsyflj-profile
## GCC version      = 7.5.0
## GSL Version      = 2.6
## OpenBlas         = OpenBLAS 0.3.9  - OpenBLAS 0.3.9 DYNAMIC_ARCH NO_AFFINITY SkylakeX MAX_THREADS=128
##   arch           = SkylakeX
##   threads        = 1
##   parallel type  = threaded
##
## Command Line Input = /home/hdenis/Programs/gemma-0.98.5-linux-static-AMD64 -bfile /nvme/disk0/lecellier_data/WGS_GBR_NC_data/Vcf_files/GWAS_files/Imputed_files/aspat_clean_all_chr_SNP_filtered_2_GQ15_0mis_imputed_maf0.05_maxgeno95_Group1_relat0.2_geno_pheno_binary -k /nvme/disk0/lecellier_data/WGS_GBR_NC_data/Vcf_files/GWAS_files/Imputed_files/aspat_clean_all_chr_SNP_filtered_2_GQ15_0mis_imputed_maf0.05_maxgeno95_Group1_relat0.2_binary.cXX.txt -lmm 1 -outdir /nvme/disk0/lecellier_data/WGS_GBR_NC_data/Assoc_studies_outputs/GEMMA/ -o Group1_GQ15_relat0.2_In_situ_mortality_binary_lmm 
##
## Date = Fri Aug  1 20:35:04 2025
##
## Summary Statistics:
## number of total individuals = 624
## number of analyzed individuals = 211
## number of covariates = 1
## number of phenotypes = 1
## number of total SNPs/var = 440918
## number of analyzed SNPs/var = 440865
## REMLE log-likelihood in the null model = -152.894
## MLE log-likelihood in the null model = -150.163
## pve estimate in the null model = 2.48451e-06
## se(pve) in the null model = 0.580001
## vg estimate in the null model = 2.51139e-06
## ve estimate in the null model = 0.251139
## beta estimate in the null model =   1.50711
## se(beta) =   0.0344997
##
## Computation Time:
## total computation time = 0.638227 min 
## computation time break down: 
##      time on eigen-decomposition = 8.90167e-05 min 
##      time on calculating UtX = 0.0119906 min 
##      time on optimization = 0.486226 min 
##``` 

Can you advise how to fix this issue ? Thank you very much !

Hugo