Implement check whether chars in sequence and component_dict match (for SequenceDesigns)

[jschepers]

Currently it's not checked for a SequenceDesign whether sequence and component_dict match:

• Case 1: Sequence contains char that does not occur in component_dict -> zeros
• Case 2: Component_dict contains char that does not occur in sequence -> SubselectDesign leads to empty design + index to add simulated responses to epoch_data is all zeros -> Extra char is ignored and there is no difference to a simulation without this char
  I would suggest implementing a check, what do you think?