Structural restrictions in local causal discovery: identifying direct causes of a target variable