transcript.add_exon(start,end) # 0-based [start,end) on genome, start < end