Which is the best way to dilucidate if a mild phenotype in a patient with primary immunodeficiency is caused by alternative splicing or somatic reversion (particularly when alternative splicing is present but it seems that something is missing to elicit a partial phenotype)?