a special probe matrix

maybe we can design a special probe matrix which can detect something like in P. 265
which consider the angle after become an alpha helix.

relation between greek key and interlock

P.184 in Introduction to Protein Structure
topo: 1 2 5 4 3 6 7

greek key: 2 5 4 3, 5 4 3 6

our 2E
(1 2)(5 4)
(7 6)3

intron position in our motifs.

check intron positions in our motifs,
ref: P. 69 P.70 in Introduction to Protein Structure.

interlock in jelly roll barrel

in P.71 and P.72 in "Introduction to Protein Structure"
it suggest a way to fold in to the structure which follow our rules.
Which, the topology is same as our 4H

about barrel

now, pdbsum has info like this in graph.
for 1neb, strand #2 should connect with strand #3,
but pdbsum missed on this one because the connection between Str#2 and Str#3
is on res#32 and res#40, but pdbsum doesn't consider #32 is part of Str#2.
and they also lost the related direction info on sheet#1 and sheet #2 (actually, sh#1 and sh#2 will be considered as one sheet in a barrel)

idea: multiple linkage clustering

http://wwwsup.scmbb.ulb.ac.be/~ocha/wwwsup1/help/sofist.paper.html

This paper use a tree like graph (multiple linkage clustering) to show align distances.
It's a pretty good idea which let you decide how many and where you want to align.
And also the way it shows the result is pretty good, it can show you the main core, the difference of those sequences.