Do common variants play a role in risk for autism? Evidence and theoretical musings