SNP-based heritability captured outside of the hg38 genome build

-------------------------------------------------------------------------------------------------------------------Source code for peforming LD score analyes comparing estimates of SNP-based heritability for hg38 vs T2T SNPsSupplementary Note 4 of Wainstchein et al. (2025)-------------------------------------------------------------------------------------------------------------------The 'archive_ldsc_T2T_hg383_comparison.tar.gz' archive contains (1) a custom C++ code to calculate LD score (named 'basic_ldsc'). The code is an a folder named basic_ldsc, which has a source code file named 'basic_ldsc.cpp', a Makefile, a eigen folder (needed for compilation) and example of result 'test.ld.score'. (2) a batch file to submt jobs to calculate LD scores using basic_ldsc (calcLDSC_with_myCode.sh) (3) an R object named 'ldsc_all_maps2.RData' containing pre-calculated LD scores for 9,025,676 (data frame named 'map2') (4) an R object named 'Inff.RData', which contains a vector named 'Inff' of inflation used to calculated effective sample size for REGENIE. Those factors are available for 29 quantitative traits. (5) an R script named 'runLDSC_v2.R' that reads 'ldsc_all_maps2.RData', 'Inff.RData' and GWAS summary statistics to perform LD score regression (LDSC) analyses. This script contains a function named 'ldsc_reg', which implements inverse probability weighting LDSC. (6) a bacth file named 'runLDSC_v2.sh' to submt jobs to calculate run 'runLDSC_v2.R' across all 29 traits. We also provide results from LDSC analyses ('ldsc_analyses_all_v2.RData') and the R script to generate Supplementary Figure 13 (FigS13.R).