How to perform weighted inference

Is there any way to use the Potential to scale the likelihood with weights while also allowing imputation or handling of missing values in y?