@RogiAshrafi @sundberglr Dependent on the type of analysis that you’re performing, you could consider using a mixed model. It can allow you to account for between-replicate variation in the response and/or the effect of the covariates. https://t.co/K4Z3GmxRy6