I know how to traverse through structures, models, residues, and atoms in a protein chain using BioPython (which is very easy).
How can I identify donor and acceptor atoms in a chain?
$\begingroup$
$\endgroup$
9
-
$\begingroup$ donor/acceptor belongs to the main chain (MC) or side chain (SC) of amino acids ?? $\endgroup$– pippo1980Oct 27, 2021 at 16:59
-
$\begingroup$ @pippo1980, side chain. My target is to measure the distance between donor and acceptor atoms. $\endgroup$– user366312Oct 27, 2021 at 17:24
-
1$\begingroup$ Ligand or hydrogen atoms of other side chains ? Just trying to put it into a perspective $\endgroup$– pippo1980Oct 27, 2021 at 17:39
-
$\begingroup$ I guess once you know coordinates of each putative donor and acceptor, you calculate the distances between all the possible pairs and take the ones under a threshold. My PhD examiner was blabbing about right geometry too … but we were talking about X-ray crystallography no way to know where the hydrogens are and a kind of middle resolution so … but remember geometry is important too $\endgroup$– pippo1980Oct 27, 2021 at 17:44
-
$\begingroup$ Here a informative link biostars.org/p/182325 this links to another post : most software simply looks for donor + acceptor pairs separated by an appropriate distance and bond angle. $\endgroup$– pippo1980Oct 27, 2021 at 17:45
|
Show 4 more comments
1 Answer
$\begingroup$
$\endgroup$
10
A hydrogen bond donor heavy atom is one that is protonated, while the acceptor has a lone pair. If you talking about amino acids, then it's simple as they are fixed.
| atom name | description | role | note |
|---|---|---|---|
N |
backbone nitrogen | donor | SS |
CA |
backbone Cα | apolar | |
C |
backbone carboxyl C | sterically hindered and electrophilic | nucleophilic attack |
O |
backbone carboxyl O | acceptor | SS |
CB etc. |
sidechain carbons | apolar | |
OXT |
C-terminal oxygen | acceptor | may be protonated especially if a membrane embedded C-terminus |
Serine's OG / Threonine's OG1 |
sidechain hydroxyl | donor | SS: may form ST turns. Fairly uncommon: may be a donor |
Asp OD1/OD2/Asn OD1 |
sidechain carboxyl/amide O | acceptor | SS: may form Asx turns |
Glu OG1/OG2/Gln OG1 |
sidechain carboxyl/amide O | acceptor | |
Lysine NZ |
sidechain amine | donor | |
Asn ND2 / Gln NG2 |
terminal admide N | donor | the direction of an Asn/Gln may be accidentally filled if manually assigned as N/O have the same density in X-ray crystals |
Water can be both a donor and an acceptor. Two charged residues form a salt bridge more so than a hydrogen bond, because the interaction is driven by the Coulomb electrostatic forces of the two heavy atoms.
-
$\begingroup$ A hydrogen bond donor heavy atom is one that is protonated, while the acceptor has a lone pair. -- Does this apply to BioPython's
Atomobject? I mean, can I test or search it using BioPython? $\endgroup$ Oct 28, 2021 at 10:49 -
$\begingroup$ Yeah nah. A heavy atom is an atom that diffracts under X-ray crystallography —not hydrogens. The
Atominstance is a graph network node representation of any atom. It may be a hydrogen —_in silico_ protonated, neutron diffraction or 1H NMR. Technically to the most pedantically punctilious level, it could be virtual (e.g. Drude particle)... $\endgroup$ Oct 28, 2021 at 10:57 -
$\begingroup$ So, can I use your supplied information and apply them on a PDB file via BioPython and extract donor and acceptor atom info from that PDB file? $\endgroup$ Oct 28, 2021 at 11:03
-
1$\begingroup$ Totally — assuming a pair is within about 3Å $\endgroup$ Oct 28, 2021 at 11:06
-
1$\begingroup$ 2.8Å is an ideal, >3.2Å is weak. Mathematically the Lenard-Jones term is used in thermodynamic simulations —in the wiki page you can see the potential plateaus quickly. Sure, the true distance will be marginally affected by the b-factors of the atoms (both anisotropic/isotropic) but in solution the distance is likely a distribution (i.e. the atoms wobbling around). $\endgroup$ Oct 28, 2021 at 17:26