admix.assoc.marginal#

admix.assoc.marginal(dset: Dataset | None = None, geno: Array | None = None, lanc: Array | None = None, pheno: ndarray | None = None, n_anc: int | None = None, cov: ndarray | None = None, method: str = 'ATT', family: str = 'linear', fast: bool = True)[source]#

Marginal association testing for one SNP at a time

Parameters:

  • dset (admix.Dataset) – data set

  • geno (da.Array) – genotype (n_snp, n_indiv, 2) matrix

  • lanc (da.Array) – local ancestry (n_snp, n_indiv, 2) matrix

  • pheno (np.ndarray) – phenotype (n_snp, )

  • n_anc (int) – number of ancestral populations, if not specified, inferred from lanc

  • cov (np.ndarray, optional) – Covariate matrix, by default None. Do NOT include 1 intercept

  • method (str, optional) – methods used for association testing, by default “ATT”, one of [“ATT”, “TRACTOR”, “JOINT”, “ADM”, “SNP1”, “ASE”, “HET”]

  • family (str, optional) – family of phenotype, by default “linear”

  • fast (bool, optional) – use fast implementation, by default True

Returns:

np.ndarray – Association p-values for each SNP being tested