heig-v1.3.0-alpha

Updates:

• Proposed a new pipeline for processing and analyzing rare variants (WES and WGS) at voxel level
• Fixed bugs in v1.2.1

Full Changelog: v1.2.1...v1.3.0-alpha

heig-v1.2.1

Updates:

• Proposed a post-GWAS screening step through --cluster
• Fixed bugs in v1.2.0

Full Changelog: v1.2.0...v1.2.1

heig-v1.2.0

Updates:

• Now LDR GWAS can be conducted by HEIG internally through --gwas
• Provided a new module --make-mt to convert bfile and VCF to hail.MatrixTable for distributed computing supported by hail
• Provided an option --skip-smoothing in FPCA step for analyzing non-imaging phenotypes
• Provided more data management and genotype data QC options
• Fixed bugs in v1.1.0

Full Changelog: v1.1.0...v1.2.0

heig-v1.1.0

Updates:

• Now reading images is a separate step --read-image
• Merged kernel smoothing and functional PCA in a single step --fpca
• Now constructing LDRs is a separate step --make-ldr, which facilities users to adjust for the number of LDRs without recomputing bases
• Supported multi-threading computation
• Improved memory usage for images with high resolution
• Changed .sumstats and .ldmatrix from pickled np.array to HDF5 file
• Changed all data from float64 to float32
• Changed the way to handle inconsistent sample size in LDR construction and LDR GWAS, which may result in ignorable difference in secondary analyses
• Not compatible with v1.0.0

Full Changelog: v1.0.0...v1.1.0

heig-v1.0.0

The initial version of HEIG v1.0.0

• Kernel smoothing and LDR generation for NIFTI, CIFTI, and FreeSurfer output
• Voxelwise GWAS for unrelated subjects
• Voxelwise heritability
• Genetic correlation between pairs of voxels
• Cross-trait genetic correlation between voxels and non-imaging phenotypes