[Biopython] Using DSSP to calculate SASA

Ahmad Abdelzaher underoath006 at gmail.com
Mon Apr 24 20:16:02 UTC 2017


Thank you, I just tried DSSP. The dictionary is a bit messy, I don't
understand its structure, and so I can't index items.

I tried the following from the DSSP class documentation, I got the
following traceback error:

>>> aa, ss, acc=dssp_dict[('A', 5)]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: tuple indices must be integers or slices, not tuple

However, putting this issue aside for now, the more serious issue is
that  It's giving somewhat different results from Pymol. Please try
the code below, and compare its results to the values obtained from
Pymol in the JSON attached below.

from Bio.PDB import PDBParser, DSSP

p = PDBParser()
structure = p.get_structure("x", "1a1x.pdb")
model = structure[0]
dssp = DSSP(model, "1a1x.pdb", dssp='dssp')
a_key = list(dssp.keys())[2]
#print(a_key)
#print(dssp[a_key])
dssp_dict=dssp_dict_from_pdb_file("1a1x.pdb")
print(dssp_dict)

To proceed with the post, O'm not sure what EXP_DSSP_RASA is exactly,
it's not mentioned in the documentation. I'm also curious about
freesasa, is there a class for it in biopython?

On Mon, Apr 24, 2017 at 9:27 PM, João Rodrigues
<j.p.g.l.m.rodrigues at gmail.com> wrote:
> Hi Ahmad,
>
> DSSP, NACCESS, and MSMS are three very different programs that give
> accessibility. NACCESS gives you per-atom and per-residue absolute and
> relative accessibilities, which are very useful in some applications. DSSP
> gives you absolute accessibilities per residue. MSMS gives you a solvent
> exposure metric (in a way).
>
> Relative means exactly that, it will divide whatever value you get from DSSP
> by a standard value for each residue. You have three different scales to
> choose from to normalize the values, whose papers are reference in the
> documentation so you should look them up and see what exactly they refer too
> and how those standard values were calculated.
>
> If you want just the SASA, without any modification, just access the values
> that the DSSP dictionary returns. If you want the relative accessibility,
> access the key 'EXP_DSSP_RASA'.
>
> NACCESS works only on Linux yes. But you have FREESASA, which should work on
> Windows with a bit of effort, is actively maintained, and has the same
> scale/references as NACCESS so it gives the same results (0.99 correlation).
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1a1x.pdb.json
Type: application/json
Size: 3604 bytes
Desc: not available
URL: <http://mailman.open-bio.org/pipermail/biopython/attachments/20170424/ccf7892e/attachment.json>


More information about the Biopython mailing list