Geometric Hashing Of Pdb Structures?

Question

[Egon Willighagen]

Joerg asked on the Blue Obelisk eXchange about geometrical hashing of PDB files. As there have not been any answers from the open source cheminformatics community yet, I am copying his question here:

Can anyone point me to geometric hashing options for PDB files? I am mainly interested in comparing two PDB files as fast as possible, and I need a comparison being more elaborate than just a Calpha comparison.

[Pawel Szczesny]

Given how many options are there: http://en.wikipedia.org/wiki/Structural_alignment_software that differ by scope of comparison it would be great if the author could clarify a bit what exactly he means when writing "more elaborate". :)

Answer 1

[Kraut]

For comparing PDB structures the first options (in order of sensitivity) that come to mind are:

• RMSD
• GDT score
• TM score

These are often used to calculate the similarity between a predicted structure and native native in experiments like CASP.

[Khader Shameer]

@Adam: Do you know about any tool/server that can calculate GDT scores ?

[Jose Manuel Duarte]

Maxcluster is a great tool to calculate both rmsds and GDT scores

Answer 2

[Chris]

Perhaps protein structure alignment is what you're looking for. Famous algorithms are e.g. DALI and CE.