JLS is a R package that implements the Joint Location-Scale testing framework that detects associated SNPs, gene-sets and pathways with main or interaction effects by leveraging complex genetic architecture (Soave et al. 2015 AJHG 1:125-138). It is recommended that the users start with the README file!

The JLS method simultaneously tests the null hypothesis of equal mean and equal variance between genotypes, by aggregating association evidence from the individual location/mean-only and scale/variance-only tests, focusing on Fisher's method of combining information (JLS-Fisher). The proposed method detects association in the presence of underlying genetic main and/or interaction effects, without specifying the interactors.

Although the work here was originally motivated by genetic association studies, application of the proposed method however is not limited to this type of data. The JLS testing framework is useful for other scientific studies where location and scale parameters are both of interest.

The previous implementation assumes there is no sample correlation or group uncertainty. A generalized Levene's scale gS and consequently the generalized jont location-scale gJLS test have been developed (Soave and Sun 2017, Biometrics). gJLS is the R package that implements the proposed method.

Contact: David Soave, david dot soave at mail dot utoronto dot ca or Lei Sun, sun at utstat dot toronto dot edu