Dynamically omit interactions from brm model

Thanks for your response. The purpose of the power calculations is to determine the sample size required to replicate the current significant result (a similar approach as described here: Power calculation for replication sample). I am using the function wp.regression() from the WebPower package to make the power calculation for each regression, and this function uses the statistic f2 which requires R^2 from the model both with and without the variable of interest. For this R^2 calculation, would you suggest I use the function brms::loo_R2() instead of brms::bayes_R2()?