Commentary: Mendelian randomization-inspired causal inference in the absence of genetic data | Synapse