As indicated in Figure 1, tandem duplications can create various problems for the above outlined alignment approach. In the following, we discuss two simple examples where duplications can confuse the segment-based alignment algorithm. Let us consider a motif that is duplicated in one or several of the input sequences S1,..., Sk. For simplicity, let us assume that our sequences do not share any significant similarity outside the motif. Moreover, we assume that the degree of similarity among all instances of the motif is roughly comparable. There are no difficulties if two sequences are to be aligned and the motif is duplicated in both sequences, i.e if one has instances  and  of the motif in sequence S1 and instances  and  of the same motif in sequence S2 as in Figure 1 (A). In such a situation, our alignment approach will correctly align  to  and  to  since, for pairwise alignment, our algorithm returns a chain of fragments with maximum total score.