Is there a way to make the model not correct by kinship?

[Hidroxiapatito]

Like I salid: Is there a way to make the model NOT correct by kinship?

I understand that setting KI = NULL makes the GAPIT function calculate a kinship matrix using the markers. And the manual says that it can be omitted by "suing BLINK method", which I don't understand what that means.

I've tried with lots of different parameters, read the source code, and even tried providing an identity matrix, but I just get the same results as if I was correcting by KI.

[p-polacek]

Hi, when using BLINK (or FarmCPU and maybe SUPER too), the kinship is derived automatically based on the pseudo-QTNs it selects.

In both FarmCPU and BLINK models, markers are tested one at a time, with pseudo QTNs added as covariates to control false positives and reduce false negatives. ... Pseudo QTNs are used to derive kinship among individuals. The model chooses a set of pseudo QTNs to derive a kinship that provides the maximum likelihood. from here

So as far as I know, you cannot run these models with no kinship or with your own kinship. To run MLM without kinship, you just use GLM.