The paper : supersecondary structure of sandwichWrite about your part : calculation and presentation 1/25/05

SCOP
I download scop’s information for all beta proteins from http://scop.mrc-lmb.cam.ac.uk/scop/data/scop.b.c.A.A.A.html
I collect scop id, pdb id, chain id, start residue number, end residue number for each scop item.

PDB
For each PDB id collected above, I download PDB files from http://www.rcsb.org/pdb/

HB info
To know hydrogen bond information, I use HBPLUS http://www.biochem.ucl.ac.uk/bsm/hbplus/home.html to calculate it.
Then I filter out HB info we don’t want (ND2, NE1, NE2, NH1, NH2, OG, OG1, OD2, OE1, etc), only leave N-O hydrogen bond.

PDBSUM
To know strands start residue number and end residue number, I use PDBSUM’s information.
For example, for 1bia, I use data in this page:
http://www.biochem.ucl.ac.uk/bsm/pdbsum/1bia/1bia_str.html
Now I know strand#1 for 1bia is from residue#51 to residue#53.
BUG: when there’s non-number character in residue number field, my program will skip that strand.

Link
I use strands number information from PDBSUM and HB information from hbplus to decide if two strands connect together.
If two strands have HB between them, they are connected. For each record in hbplus, I check which strands they belongs to using PDBSUM’s info, and output strands numbers in format “#A=#B #C D #E F”. It means on strand number A residue number C which is residue D has an hydrogen bond with strand number B at residue number E which is residue F.
For example: “1=4 7 ILE 34 THR”

Graph
To get graph represent strands connectivity, I use graphviz http://www.graphviz.org/
I use the link information from above section and translate the information into graphviz’s format and use graphviz to generate GIF graphs for each protein on scop.

Distances
For calculating Strand-Strand distance, I take every possible combinations of one residue on two strands each, and calculate the distances use PDB file’s xyz information, then choose a smallest one as Strand to Strand distance.

Secondary Structure
I use Promotif to calculate Secondary Structure. And cut the last section of Promotif’s output, rearrange the format, then save it in both csv format and text format.

HTML
I use scop’s html http://scop.mrc-lmb.cam.ac.uk/scop/data/scop.b.c.A.A.A.html
And insert our data for each scop item after the scop id.

01/29/05 Findings:
1rl2A
in PDBSUM:
10 170 174
11 180 184
in HB: A0171-VAL N A0183-ILE O
10=11為何沒出現在LINK裡?



for 1rl2, TOPS 比較利害

for 1dj7, #2 #5 比較接近 但在TOPS比較遠 可見他沒有表現出遠近關係
而且 他看不出 2-3 和 5-6 是無關(沒有連在一起)的

01/31/05
find structures with parallel strands
lot’s of parallel strands, ex:1bww 10,11 with 2. 1iqd 9,10 with 2

find structures with more than two sheets
too hard. Have to do it manually.