[ Back to geometry
and molecular biology ] 
The HTH motif we have detected here is quite important: it implies16% and 25% of the two proteins (17 amino acids). To detect motifs with less than 10 to 15 amino acids, we would have to discard all matches between secondary structure elements, such as alpha helices.
The use of frames instead of just points allows the algorithm to discard several false matches of amino acids (4 in the HTH case). Moreover, it makes the algorithm robust with respect to parameter that tune the quality of the matches (which is the uncertainty that we fix on frames on input).
However, designing a 6D hash table with uncertainty on rigid transformations (the invariant space) is far from simple and our current implementation requires too many parameters and is not robust. Solving this problem would give a fast, efficient, and fully automated algorithm for substructure matching.
[ Back to geometry
and molecular biology ]