In this subsection, we discuss which of the 5 alignment classes have the best DIALIGN score.
For the two most conserved alignment classes, we analyzed the persistence of the sequence similarity in more distantly related species.
The DIALIGN alignment was also shown to be stable (see Fig. 7) whereas a variety of different alignment classes showed up in the CLUSTAL W alignments, one example being also shown in Fig. 7.
And, indeed, we could relate distinct types of tree topologies to distinct alignment classes which were defined in terms of the relative position of the – A – S – B – motifs.
For the genome alignment classes, NM indicates reads that could not be mapped to the genome, R* reads that were mapped to repeat areas and U0-U2, reads that were mapped with either no, one or two mismatches to the genome.
To compute DIALIGN scores for all alignment classes, the xfr option described in 2 was used: We excluded the possibility of aligning the subwords S1 and B1 of S22 with subwords of any of the other sequences aligned with these subwords in the class-2 or class-4 alignments.
In doing so, we classified these alignments by the relative positions of the motifs A, S, and B. This classification led to the 5 alignment classes 1a, 1b, 2, 3, and 4 shown in Fig. 4 one of which, class 1a, was observed only once, viz., for ⅅ14.
Indicated from top to bottom are the total number of input sequence reads, sequence reads that were shorter than 15 nt after 3' adapter truncation, sequences accepted for further analysis after 3' adapter truncation, genome alignment classes and the number of sequences annotated to non-coding RNA transcripts.
We performed two types of alignments, class 4 HDACs with HDACs of other classes and class 4 HDACs alone.
On the basis of sequence and structural alignments, class I cyts c have further been partitioned into 16 different subclasses.
Among these alignments, alignments in class 3 look biologically most meaningful since it implies only one duplication event.
