solareclipser

solareclipser is an R package for SOLAR-Eclipse. It ports SOLAR’s Fast Permutation Heritability Inference (FPHI) to R:

• load an empirical kinship pedigree (IDA, IDB, KIN) and a phenotype file
• adjust a trait for covariates, in SOLAR’s covariate syntax (age^1,2#sex)
• estimate its heritability (h2r) with standard errors and a likelihood ratio test, returned as data frames
• optionally write the results as csv, tsv, json, yaml, or the summary block of the original SOLAR fphi command

Results are checked against SOLAR 9.0.1 in the package tests.

Installation

base R installation (building from source needs a C++ and a Fortran compiler, e.g. gfortran, and the zlib headers, e.g. zlib1g-dev on Debian/Ubuntu):

1. Download the tarball from the releases page

2. Install the package with the following command (replace the path with the actual path to the downloaded file):

   install.packages("path/to/package.tar.gz", repos = NULL, type = "source")

Example

The package bundles a Human Connectome Project kinship pedigree and white matter phenotypes. The functions take file names, as SOLAR does, so write them to CSV first:

library(solareclipser)

dir <- tempfile("fphi_")
dir.create(dir)
write.csv(pedigree, file.path(dir, "ped.csv"), row.names = FALSE, quote = FALSE)
write.csv(phenotypes, file.path(dir, "phen.csv"), row.names = FALSE, quote = FALSE)

solar_load_pedigree(file.path(dir, "ped.csv"), threshold = 0, output_dir = dir)
solar_load_phenotype(file.path(dir, "phen.csv"))
solar_select_trait("CC")
res <- solar_run_fphi(file.path(dir, "CC"))
res$results
#>   trait      h2r         se    loglik sporadic_loglik      p_value n_subjects
#> 1    CC 0.934443 0.01040862 -332.2292       -487.5209 8.148458e-70        999
res$parameters
#>   parameter      value         se
#> 1      mean 0.00404806 0.11845822
#> 2        e2 0.06555697 0.01040862
#> 3       h2r 0.93444303 0.01040862
#> 4        sd 1.04560387 0.02609633

To adjust for covariates, select them before running FPHI, e.g. solar_select_covariates("age^1,2#sex") for age, age^2, sex and their interactions (the bundled phenotypes have no covariate columns). To also get the results as text, pass format, e.g. solar_run_fphi(base, format = "solar") for output that can be compared with the original SOLAR.