Inf and NA in the results

[Kai6662]

Hi,

My results have amounts of Inf and NA. Why did it happen like that?

Best regards,
Kai

[Rick-Chen-PKU]

I have the same problem and hope the developers pay attention to this issue

[zh-zhang1984]

I have the same issues in running correlation between BMI and HDL:

library(HDL)
HDL_exp_dat <- HDL_exp %>% 
  mutate( N = 403943) %>% 
  dplyr::select(
    SNP,N,
    A1 = effect_allele.exposure,A2 = other_allele.exposure,
    N,
    b = beta.exposure, se = se.exposure,
    Z = beta.exposure/se.exposure
  )
BMI_exp_dat <- BMI_exp %>% 
  mutate(N = 681275) %>% 
  dplyr::select(
    SNP,
    A1 = effect_allele.exposure,A2 = other_allele.exposure,
    N,
    b = beta.exposure, se = se.exposure,
    Z = beta.exposure/se.exposure
  )
LD.path <- "/Users/zhangzhongheng/Documents/2022/GWAS_sepsis/UKB_imputed_SVD_eigen99_extraction"
LD.path <- "/Users/zhangzhongheng/Documents/2022/GWAS_sepsis/UKB_imputed_hapmap2_SVD_eigen99_extraction"
res.HDL <- HDL.rg(HDL_exp_dat, BMI_exp_dat, LD.path)
res.HDL
$rg
    
Inf 

$rg.se
[1] NaN

$P
    
NaN 

$estimates.df
                        Estimate           se
Heritability_1      0.000000e+00 0.000000e+00
Heritability_2      0.000000e+00 0.000000e+00
Genetic_Covariance  1.224554e-07 7.707554e-08
Genetic_Correlation          Inf          NaN

$eigen.use
[1] 0.9

The same two GWAS, when running with lds, is reporting a negative correlation.