Question: Is there any easy way to get ASA/RSA for the whole PDB?
0
gravatar for ajingnk
5.3 years ago by
ajingnk130
United States
ajingnk130 wrote:

I want to get the accessible surface area for each residue of a protein structure. I need to do that for biological assembly and remove ligands. I hope this could cover all proteins in UniProt. However, I don't have an efficient way to remove ligands. I have data from SIFTS(http://www.ebi.ac.uk/pdbe/docs/sifts/), which maps PDB residue to UniProt residue. What I am thinking is:

1) I calculate ASA/RSA for all PDBs(Biological Assembly)

2) I take the maximum ASA for all corresponding residues, and assign the maximum ASA to the residue on UniProt

Is this a right way to do it? Because, here I assume, I will always find a PDB without any ligand for each protein in UniProt.

I am thinking to use DSSP or NACCESS. I hope these two tools can deal with water and other molecule correctly.

 

Thanks

asa protein pdb • 1.7k views
ADD COMMENTlink modified 5.3 years ago by vavrusa80 • written 5.3 years ago by ajingnk130
0
gravatar for vavrusa
5.3 years ago by
vavrusa80
France
vavrusa80 wrote:

I'm not sure if I understand the question right, but the ligand residues are stored as HETATM, just grep them out.

$ grep -vi HETATM 1avx.pdb > 1avx_nohet.pdb

Note that this also removes non-standard residues, substrate, solvents etc. which you might not want in the PDB anyway for the SA calculation.

ADD COMMENTlink written 5.3 years ago by vavrusa80

The problem is that HETATM can be part of protein. Because, modified residues will also be assigned as HETATM. ION is HETATM which may be part of protein also. And, atoms which are not HETATM may not be part of the protein, like binding peptides.

ADD REPLYlink written 5.3 years ago by ajingnk130

Makes sense. If you don't know how the composition of the ligands or have PDBs with modified residues or the ligands are part of protein chains, then I'm not sure if I can think of an easy way. What you propose, reconstructing the protein ASA sans the ligand, may work. You can do it the other way round, filter out the ligands based on the ligand database search, but I suppose there's a risk of filtering legitimate parts of the protein. Neither is "easy" computation-wise.

ADD REPLYlink written 5.3 years ago by vavrusa80
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1104 users visited in the last hour