search

mglev1n / ldscr

LD Score Regression in R
https://mglev1n.github.io/ldscr/
Other
31 stars 3 forks source link

what is 'S_Stand' ? #18

Closed Libertas2017 closed 2 months ago

Libertas2017 commented 2 months ago

! Your genetic covariance matrix includes traits estimated to have a negative heritability. Error: object 'S_Stand' not foundoking'

library(ldscr) rg_res <- ldsc_rg(

  • munged_sumstats = list(
  • "COPD" = COPD,
  • "Smoking" = Smoking
  • ),
  • ancestry = "EUR"
  • ) ℹ No sample prevalence data provided. Estimating heritabilities on the observed scale. ✔ Using EUR reference from Pan-UKB [51ms] ✔ Reading LD Scores [252ms]
    ✔ Reading weights [159ms]
    ✔ Reading M [39ms]
    ✔ Reading summary statistics for 'COPD' from dataframe [16ms] ℹ 1082885/11831889 SNPs remain after merging 'COPD' with LD-score files ℹ Removed 0 SNPs with Chi^2 > 455.449 from 'COPD'; 1082885 SNPs remain ✔ Merging 'COPD' with LD-score files [9.7s] ✔ Reading summary statistics for 'Smoking' from dataframe [14ms] ℹ 1093765/8683163 SNPs remain after merging 'Smoking' with LD-score files ℹ Removed 0 SNPs with Chi^2 > 80 from 'Smoking'; 1093765 SNPs remain ✔ Merging 'Smoking' with LD-score files [5.6s] ✔ Reading summary statistics [15.4s] ✔ Estimating heritability for 'COPD' [144ms] ✔ Estimating genetic covariance for for 'COPD' and 'Smoking' [1.5s] ! Your genetic covariance matrix includes traits estimated to have a negative heritability. Error: object 'S_Stand' not foundoking' ✖ Estimating heritability for 'Smoking' [469ms]
mglev1n commented 2 months ago

That error suggests there is a problem with your GWAS summary statistics (eg. using a small, underpowered GWAS without significant trait heritability), and comes from the GenomicSEM package (https://github.com/GenomicSEM/GenomicSEM/) which ldscr leverages for calculating heritability and genetic correlations.