Fig. 1

Alignment of the consensus sequences of five Alu sub-families. Dashes indicate gaps introduced so as to maximise alignment. Nucleotides identical between all sequences are indicated by asterisks. Pathological Alu insertions (including 28 simple ones and three associated with large genomic deletions) that are informative with respect to starting position in their respective Alu sub-family consensus sequences, are positioned accordingly in the aligned sequences. Note that the sub-family of the shortest Alu insert, which comprises CGTCTC plus A₄₀and is associated with the Δ1444 bp in the SERPINC1 gene (Beauchamp et al. 2000; Chen et al. 2005), could not be assigned. Shaded arrows indicate either entries (underlined) that can be alternatively annotated as full-length Alu inserts or those that are not informative with respect to the ‘microhomology’ question (refer to Supplementary Table S1 for details). Note that (i) microhomology existing between the top strand’s 5′ overhang and the sequence that lies 5′ to the truncation position in the Alu consensus sequence was identified in the same way as for the Alu insertion polymorphisms (see second section of the text) and (ii) only Alu inserts with starting position 6 or greater were regarded as 5′-truncated in accordance with Zingler et al. (2005)