H-bonded features: First H bonds are defined, based on them, turns and bridges are defined

1. Dictionary of Secondary Structure of Proteins (DSSP) -Pattern Recognition H-based and Geometrical features

2. DSSP algorithm is the standard method for assigning secondary structure to the amino acids of a protein, given the atomic-resolution coordinates of the protein DSSP recognizes eight types of secondary structure (each identified by its own symbol), depending on the pattern of hydrogen bonds These eight types are usually grouped into three larger classes: helix (G, H and I), strand (E and B) and loop (all others - S,L,T)

3. H-bonded features: First H bonds are defined, based on them, turns and bridges are defined Then Alpha-Helices and Beta-Ladders including helical kinks and Beta-bulges are defined Geometrically defined features: Bends Chirality SS bonds Solvent exposure Each structural feature is defined independently of the others and structural overlaps are resolved by defining a secondary structure summary that assigns a single state to each residue.

4. Hydrogen bonds: n-turn: CO(i) and NH(i+n) where n=3,4,5 Bridge: Two nonoverlapping stretches of three residues (i-1,1,1+1) and (j-1,j,j+1) Hydrogen bonded structure

5. Co-operative H-bond patterns : Helices Minimal helix defined by two consecutive n-turns Beta ladders and sheets

6. Bend : regions with high curvature Chirality : + sign if - sign if Geometrical structure

7. The secondary structure assignments can only be as accurate as the coordinates on which they are based. It is important to be aware of the state of resolution and refinement of each structure. One can conlcude that the criteria for H-bonded secondary structure realtively strict, inspite of a generous cutoff in the H-bonding energy. For high resolution data sets, these assignments are more accureate than those for lower-resolution data. Accuracy of H-bond and Secondary Structure Assignments

8. What are the preferred lengths of secondary structure segments? The length distributions fall off almost monotonically with increasing length upto a max segment length of about 30 A There appear to be no statistically signinficant peaks either for integral number of helical repeats or for typical domain sizes with the possible exception of four-residue parallel beta-ladders characteristicof the // folding unit. The protein folding follows random polymer statistics aroximately in that long segments are statistically less likely than short ones. Secondary structure size b