Pretty neat stuff. The method used for decimating representations is pretty awesome. Seeing the aligned routines makes me think that quite a few of the bioinfor

Pretty neat stuff. The method used for decimating representations is pretty awesome. Seeing the aligned routines makes me think that quite a few of the bioinformatics algorithms could be useful for such an analysis. The two that spring to mind would are the MCL clustering algorithm that could be applied quite easily to the similarity matrix. As a more academic endeavor, it'd also be interesting to see what kinds of differences in similarity you'd get by applying Needleman-Wunsch. http://www.micans.org/mcl/ http://www.micans.org/mcl/ http://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm http://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algori...