cellink.at.GWAS

class cellink.at.GWAS(Y, F=None)

Linear model for univariate association testing between 1 phenotypes and S inputs (1`x`S tests)

Methods table

getBetaSNP()	get effect size SNPs
getBetaSNPste()	get standard errors on betas
getLRT()	get lik ratio test statistics
getPv()	Get pvalues
test_association(G)	Test association between phenotype and genotype matrix.

Methods

GWAS.getBetaSNP()

get effect size SNPs

Return type:

ndarray

Returns:

-beta (ndarray)

GWAS.getBetaSNPste()

get standard errors on betas

Return type:

ndarray

Returns:

-beta_ste (ndarray)

GWAS.getLRT()

get lik ratio test statistics

Return type:

ndarray

Returns:

-lrt (ndarray)

GWAS.getPv()

Get pvalues

Return type:

ndarray

Returns:

-pv (ndarray)

GWAS.test_association(G)

Test association between phenotype and genotype matrix.

Each column of G is tested independently from the others. The test is performed using the likelihood ratio test (LRT) statistic. The LRT statistic is computed as: .. math:

LRT = -df * log( marginal likelihood under H1 / marginal likelihood under H0 )

where s2 is the variance of the residuals of the model with the covariate and s20 is the variance of the residuals of the null model. Uses the Woodbury Matrix Identity to invert the matrix in the LRT statistic. Fit genotypes one-by-one.

Parameters:

G ((N, S) ndarray) – inputs

Return type:

None