Skip to content

Commit 40620d2

Browse files
slipchenkoslipchenko
committed
restructure
1 parent 0c002ba commit 40620d2

142 files changed

Lines changed: 221 additions & 23509 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎Untitled‎

Lines changed: 0 additions & 9 deletions
This file was deleted.

Makefile renamed to doc/Makefile

File renamed without changes.

‎doc/README.md‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+

applied_efp.rst renamed to doc/applied_efp.rst

Lines changed: 52 additions & 53 deletions
Original file line numberDiff line numberDiff line change
@@ -7,99 +7,98 @@ FMO: Applied flexible EFP
77
Overview
88
--------
99

10-
As is the case with many photoactive proteins,computational methods struggle to reproduce
11-
experimental spectra for the Fenna-Matthews-Olson complex (FMO). Work by
12-
`Kim et al <https://pubs.acs.org/doi/full/10.1021/acs.jpclett.9b03486>`_ shows that
13-
flexible QM/EFP can be applied to FMO to correctly generate computational results in
14-
quantitative agreement to experimental spectra.
15-
16-
The key to applying EFP to your system is to carefully define the active site and EFP region.
17-
FMO is a trimeric protein with eight bacteriochloropyll a (BChl) pigments in each monomer.
18-
FMO completes energy transfer via excitonic couplings across these eight BChls. A summary
19-
of the complete workflow that was performed is the following: 1) molecular dynamics (MD)
20-
simiulations of the FMO protein in water and counter ions, 2) QM/MM (not EFP) geometry
21-
optimization of *each* active site (active sites consist of one BChl pigment and
22-
typically 3 H-bonding amino acids), and 3) flex-EFP excited state energy calculations of
10+
As is the case with many photoactive proteins,computational methods struggle to reproduce
11+
experimental spectra for the Fenna-Matthews-Olson complex (FMO). Work by
12+
`Kim et al <https://pubs.acs.org/doi/full/10.1021/acs.jpclett.9b03486>`_ shows that
13+
flexible QM/EFP can be applied to FMO to correctly generate computational results in
14+
quantitative agreement to experimental spectra.
15+
16+
The key to applying EFP to your system is to carefully define the active site and EFP region.
17+
FMO is a trimeric protein with eight bacteriochloropyll a (BChl) pigments in each monomer.
18+
FMO completes energy transfer via excitonic couplings across these eight BChls. A summary
19+
of the complete workflow that was performed is the following: 1) molecular dynamics (MD)
20+
simiulations of the FMO protein in water and counter ions, 2) QM/MM (not EFP) geometry
21+
optimization of *each* active site (active sites consist of one BChl pigment and
22+
typically 3 H-bonding amino acids), and 3) flex-EFP excited state energy calculations of
2323
each pigment.
2424

25-
In the case of FMO, these steps must be repeated on several snapshots from MD to account
26-
for variation in the resting state of the structure, and the QM region must be defined
27-
carefully in both the QM/MM and flex-EFP stages. It might not be universally true that one
28-
must perform QM/MM geometry optimization. This page is a walkthrough for the flex-EFP procedure
29-
only. Molecular dynamics and QM/MM optimizations are assumed to be complete for your
30-
system prior to these steps.
25+
In the case of FMO, these steps must be repeated on several snapshots from MD to account
26+
for variation in the resting state of the structure, and the QM region must be defined
27+
carefully in both the QM/MM and flex-EFP stages. It might not be universally true that one
28+
must perform QM/MM geometry optimization. This page is a walkthrough for the flex-EFP procedure
29+
only. Molecular dynamics and QM/MM optimizations are assumed to be complete for your
30+
system prior to these steps.
3131

32-
.. image:: images/FMO_trimer_BCLs.bmp
32+
.. image:: ../images/FMO_trimer_BCLs.bmp
3333
:width: 350
34-
35-
.. image:: images/FMO_mon_pigs.bmp
34+
35+
.. image:: ../images/FMO_mon_pigs.bmp
3636
:width: 400
3737

3838
You will need a structure file (.g96) and topology information (.top, for atom charges). In this specific case,
39-
a structure file is extracted from a GROMACS molecular dynamics trajectory and all water molecules more than 15 angstroms from
40-
the protein's surface have been removed. For a chlorophyll-containing protein, you will likely want to optimize the geometry
41-
of each active chlorophyl molecule (with very close amino acids/water molecules) separately with more standard QM/MM approaches
42-
before proceeding with EFP calculations on the optimized geometry. For this example, the first BChl, residue number 359,
39+
a structure file is extracted from a GROMACS molecular dynamics trajectory and all water molecules more than 15 angstroms from
40+
the protein's surface have been removed. For a chlorophyll-containing protein, you will likely want to optimize the geometry
41+
of each active chlorophyl molecule (with very close amino acids/water molecules) separately with more standard QM/MM approaches
42+
before proceeding with EFP calculations on the optimized geometry. For this example, the first BChl, residue number 359,
4343
has been optimized and will be the QM region for the EFP calcuation.
4444

45-
.. image:: images/fmo_waters15a.bmp
45+
.. image:: ../images/fmo_waters15a.bmp
4646
:width: 400
4747

48-
First, an EFP region must be defined. Every amino acid, (non QM) BChl, and water molecule containing an
49-
atom within 15 angstroms of the QM BChl headring.
48+
First, an EFP region must be defined. Every amino acid, (non QM) BChl, and water molecule containing an
49+
atom within 15 angstroms of the QM BChl headring.
5050

51-
The headring is defined by atomnames: MG CHA CHB HB CHC HC CHD HD NA C1A
52-
C2A H2A C3A H3A C4A CMA HMA1 HMA2 HMA3 NB C1B C2B C3B C4B CMB HMB1 HMB2 HMB3 CAB OBB CBB HBB1 HBB2 HBB3 NC C1C C2C H2C C3C
53-
H3C C4C CMC HMC1 HMC2 HMC3 CAC HAC1 HAC2 CBC HBC1 HBC2 HBC3 ND C1D C2D C3D C4D CMD HMD1 HMD2 HMD3 CAD OBD CBD HBD CGD O1D O2D
51+
The headring is defined by atomnames: MG CHA CHB HB CHC HC CHD HD NA C1A
52+
C2A H2A C3A H3A C4A CMA HMA1 HMA2 HMA3 NB C1B C2B C3B C4B CMB HMB1 HMB2 HMB3 CAB OBB CBB HBB1 HBB2 HBB3 NC C1C C2C H2C C3C
53+
H3C C4C CMC HMC1 HMC2 HMC3 CAC HAC1 HAC2 CBC HBC1 HBC2 HBC3 ND C1D C2D C3D C4D CMD HMD1 HMD2 HMD3 CAD OBD CBD HBD CGD O1D O2D
5454
CED HED1 HED2 HED3
5555

5656
The headring surrounded by EFP region looks like this:
5757

58-
.. image:: images/tester.bmp
58+
.. image:: ../images/tester.bmp
5959
:width: 400
6060

61-
EFP is, of course, a fragmentation method. The protein residues within the 15 angstrom cutoff will be expressed individually.
62-
Because amino acids are a continuous chain, we will need to break each residue into its own fragment. Chemically, we would like
63-
to divide each residue by the C-C backbone bond, however, standard PDB listing convention divides residues by the C-N
64-
bond. To correct this, 'C' and 'O' atom names should be included in the following aminoc acid. This way the 'C' and 'CA'
61+
EFP is, of course, a fragmentation method. The protein residues within the 15 angstrom cutoff will be expressed individually.
62+
Because amino acids are a continuous chain, we will need to break each residue into its own fragment. Chemically, we would like
63+
to divide each residue by the C-C backbone bond, however, standard PDB listing convention divides residues by the C-N
64+
bond. To correct this, 'C' and 'O' atom names should be included in the following aminoc acid. This way the 'C' and 'CA'
6565
(carbonyl carbon and alpha carbon respectively) bond is the division between bonded fragments.
6666
See the example below:
6767

68-
.. image:: images/pdb_67_col.bmp
68+
.. image:: ../images/pdb_67_col.bmp
6969
:width: 400
7070

7171
For EFP, we would like these two fragments to look like this:
7272

73-
.. image:: images/efp_67_col.bmp
73+
.. image:: ../images/efp_67_col.bmp
7474
:width: 400
7575

7676
The desired atoms are contained in the structure file, but they do not completely 'agree' with the amino acid numbering.
77-
Below is a snippet from the structure file with the desired EFP fragment 8 highlighted. Note that atom names 'C' and 'O'
77+
Below is a snippet from the structure file with the desired EFP fragment 8 highlighted. Note that atom names 'C' and 'O'
7878
have to be included in the following fragment.
7979

80-
.. literalinclude:: ./examples/flex-EFP/1.Prepare_Structure/bchl359-50028.g96
80+
.. literalinclude:: ../examples/flex-EFP/1.Prepare_Structure/bchl359-50028.g96
8181
:linenos:
8282
:lines: 79-101
8383
:emphasize-lines: 10-21
8484

85-
Next, the BChl molecules are closer than the 15 angstrom cutoff, so they also appear in the EFP region. It is more cost efficient
85+
Next, the BChl molecules are closer than the 15 angstrom cutoff, so they also appear in the EFP region. It is more cost efficient
8686
to treat BChl fragments as separate head and tail groups as is shown below:
8787

88-
.. images/efp_headtail.bmp
88+
.. ../images/efp_headtail.bmp
8989
:width: 400
90-
91-
In the case of both amino acid and BChl fragments, we have at least one broken bond; we cannot simply compute a fragment that is
92-
missing an atomic bond. To solve this, we will introduce virtual hydrogen atoms to 'cap' the broken bonds. Non terminal
93-
amino acid fragments will have two virtual atoms. The first is between the alpha carbon of the previous residue and the
94-
carbonyl carbon of the current residue; the other is similarly between the alpha carbon of the current residue and the carbonyl carbon
95-
of the following residue. The BChl fragments are split between atoms 'C2A' and 'CAA.' One virtual atom will be added to both
96-
head and tail fragments between these atoms. Virtual atoms are added along the vector of the broken bonds with the distance changed
90+
91+
In the case of both amino acid and BChl fragments, we have at least one broken bond; we cannot simply compute a fragment that is
92+
missing an atomic bond. To solve this, we will introduce virtual hydrogen atoms to 'cap' the broken bonds. Non terminal
93+
amino acid fragments will have two virtual atoms. The first is between the alpha carbon of the previous residue and the
94+
carbonyl carbon of the current residue; the other is similarly between the alpha carbon of the current residue and the carbonyl carbon
95+
of the following residue. The BChl fragments are split between atoms 'C2A' and 'CAA.' One virtual atom will be added to both
96+
head and tail fragments between these atoms. Virtual atoms are added along the vector of the broken bonds with the distance changed
9797
to the C-H equilibrium bond distance, 1.09 angstroms.
9898

99-
.. images/efp_bothvirt.bmp
99+
.. ../images/efp_bothvirt.bmp
100100
101-
Once the system is properly fragmented, we can finally run EFP calculations in the precense of the polarizable, solvatochromic environment.
101+
Once the system is properly fragmented, we can finally run EFP calculations in the precense of the polarizable, solvatochromic environment.
102102

103103
EFP Workflow
104104
------------
105-

bioefp.rst renamed to doc/bioefp.rst

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ BioEFP
77
BioEFP is a scheme to split a macromolecule into fragments.
88
Proteins can be split along C-Ca bonds, as shown below.
99

10-
.. image:: images/BioEFP.jpg
10+
.. image:: ../images/BioEFP.jpg
1111
:width: 200
1212

1313
The model is described `here <https://pubs.acs.org/doi/abs/10.1021/acs.jpcb.6b04166>`_.

conf.py renamed to doc/conf.py

File renamed without changes.

efp_calcs.rst renamed to doc/efp_calcs.rst

File renamed without changes.

‎doc/efp_parameter_databases.rst‎

Lines changed: 130 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,130 @@
1+
.. _efp_parameter_databases:
2+
3+
***********************
4+
EFP parameter databases
5+
***********************
6+
7+
S22 Parameters:
8+
^^^^^^^^^^^^^^^
9+
Geometries from `doi.org/10.1039/B600027D <http://dx.doi.org/10.1039/b600027d>`_
10+
11+
- .. raw:: html
12+
13+
<a href="examples/parameters/2aminopyridine.efp" target="_blank">2aminopyridine.efp</a>
14+
- .. raw:: html
15+
16+
<a href="examples/parameters/adenine-stack.efp" target="_blank">adenine-stack.efp</a>
17+
- .. raw:: html
18+
19+
<a href="examples/parameters/adenine-wc.efp" target="_blank">adenine-wc.efp</a>
20+
- .. raw:: html
21+
22+
<a href="examples/parameters/ethene.efp" target="_blank">ethene.efp</a>
23+
- .. raw:: html
24+
25+
<a href="examples/parameters/ethyne.efp" target="_blank">ethyne.efp</a>
26+
- .. raw:: html
27+
28+
<a href="examples/parameters/formicacid.efp" target="_blank">formicacid.efp</a>
29+
- .. raw:: html
30+
31+
<a href="examples/parameters/hydrogencyanide.efp" target="_blank">hydrogencyanide.efp</a>
32+
- .. raw:: html
33+
34+
<a href="examples/parameters/indole.efp" target="_blank">indole.efp</a>
35+
- .. raw:: html
36+
37+
<a href="examples/parameters/pyrazine.efp" target="_blank">pyrazine.efp</a>
38+
- .. raw:: html
39+
40+
<a href="examples/parameters/pyridone.efp" target="_blank">pyridone.efp</a>
41+
- .. raw:: html
42+
43+
<a href="examples/parameters/thymine-stack.efp" target="_blank">thymine-stack.efp</a>
44+
- .. raw:: html
45+
46+
<a href="examples/parameters/thymine-wc.efp" target="_blank">thymine-wc.efp</a>
47+
- .. raw:: html
48+
49+
<a href="examples/parameters/uracil.efp" target="_blank">uracil.efp</a>
50+
51+
S66 Parameters:
52+
^^^^^^^^^^^^^^^
53+
Geometries from `dx.doi.org/10.1021/ct200673a <http://dx.doi.org/10.1021/ct200673a>`_
54+
55+
- .. raw:: html
56+
57+
<a href="examples/parameters/acetamide-gp.efp" target="_blank">acetamide-gp.efp</a>
58+
- .. raw:: html
59+
60+
<a href="examples/parameters/acetamide-hb.efp" target="_blank">acetamide-hb.efp</a>
61+
- .. raw:: html
62+
63+
<a href="examples/parameters/aceticacid-gp.efp" target="_blank">aceticacid-gp.efp</a>
64+
- .. raw:: html
65+
66+
<a href="examples/parameters/aceticacid-hb.efp" target="_blank">aceticacid-hb.efp</a>
67+
- .. raw:: html
68+
69+
<a href="examples/parameters/ammonia.efp" target="_blank">ammonia.efp</a>
70+
- .. raw:: html
71+
72+
<a href="examples/parameters/benzene.efp" target="_blank">benzene.efp</a>
73+
- .. raw:: html
74+
75+
<a href="examples/parameters/cyclopentane.efp" target="_blank">cyclopentane.efp</a>
76+
- .. raw:: html
77+
78+
<a href="examples/parameters/formamide.efp" target="_blank">formamide.efp</a>
79+
- .. raw:: html
80+
81+
<a href="examples/parameters/methane.efp" target="_blank">methane.efp</a>
82+
- .. raw:: html
83+
84+
<a href="examples/parameters/methanol.efp" target="_blank">methanol.efp</a>
85+
- .. raw:: html
86+
87+
<a href="examples/parameters/methylamine.efp" target="_blank">methylamine.efp</a>
88+
- .. raw:: html
89+
90+
<a href="examples/parameters/neopentane.efp" target="_blank">neopentane.efp</a>
91+
- .. raw:: html
92+
93+
<a href="examples/parameters/nmethylacetamide.efp" target="_blank">nmethylacetamide.efp</a>
94+
- .. raw:: html
95+
96+
<a href="examples/parameters/pentane.efp" target="_blank">pentane.efp</a>
97+
- .. raw:: html
98+
99+
<a href="examples/parameters/phenol.efp" target="_blank">phenol.efp</a>
100+
- .. raw:: html
101+
102+
<a href="examples/parameters/pyridine.efp" target="_blank">pyridine.efp</a>
103+
- .. raw:: html
104+
105+
<a href="examples/parameters/uracil-gp.efp" target="_blank">uracil-gp.efp</a>
106+
- .. raw:: html
107+
108+
<a href="examples/parameters/water.efp" target="_blank">water.efp</a>
109+
110+
Amino acid Parameters:
111+
^^^^^^^^^^^^^^^^^^^^^^
112+
- .. raw:: html
113+
114+
Parameters in <a href="examples/parameters/1.6-31g-d" target="_blank">6-31G(d)</a> basis
115+
116+
- .. raw:: html
117+
118+
Parameters in a hybrid <a href="examples/parameters/6.6-31g-d_6-31+xg-3df-2p" target="_blank">6-31G(d)/6-31-+G(3df,2p)</a> basis
119+
120+
- .. raw:: html
121+
122+
Parameters in a hybrid <a href="examples/parameters/7.6-31g-d_6-311++g-3df-2p" target="_blank">6-31G(d)/6-311++G(3df,2p)</a> basis
123+
124+
The fragment parameters in FlexibleEFP are adjusted based on translations and rotations of local coordinate frames associated with fragment atoms to accommodate different fragment geometries. A parameter database for standard amino acids was developed to automate flexible EFP simulations in proteins using cryptochrome 1 protein (Arabidopsis thaliana's, Cry1At, PDB: 1U3D64). The applicability of flexible EFP was demonstrated in large-scale protein simulations, where binding energies, as well as vertical electron ionization and electron attachment energies of a lumiflavin chromophore in the cryptochrome 1 protein were computed. The results showed that flexible EFP closely agrees with the standard EFP procedure but with a significant reduction in computational cost. Twenty-five protein conformations from molecular dynamics trajectories of cryptochrome 1 were fragmented into individual amino acid (AA) fragments along Cα−C bonds, yielding 12,125 BioEFP AA fragments. To improve accuracy, each AA fragment was further split along the Cα−Cβ bond into a backbone group and a side-chain group, and disulfide bridges were fragmented along the S−S bond. To recombine the backbone and side-chain fragments into a complete AA moiety, parameters at a bond midpoint and the LMO centroid between Cα and Cβ were excluded to ensure the stability of the polarization self-consistent procedure. For further details please refer to:
125+
126+
- Database of fragments describing amino acid residues and peptide backbone groups in different geometries `Flexible EFP paper <https://doi.org/10.1021/acs.jctc.0c00758>`_.
127+
- Strategy of rotating and shifting the parameters <https://doi.org/10.1021/acs.jctc.0c00758>`_.
128+
- Script that will do it for you `parameter rotation script <https://github.com/libefp2/libefp/blob/master/tools/Flexible_V5.py>`_.
129+
- Splitting of the protein into amino acid fragments and matching the parameters to specific geometries of the fragments with BioEFP and FlexEFP tutorials is described in :ref:`bioefp`.
130+
- Note: Parameters from the database should be matched to the geometry of your system.

‎doc/efpmd.rst‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
.. _libefp:
2+
3+
******
4+
EFPMD
5+
******
6+
7+
Documentation of `efpmd` driver `efpmd <https://github.com/libefp2/libefp/tree/master/efpmd>`_ or
8+
:download:`EFPMD manual <../examples/efpmd_manual.pdf>`

flexible_efp.rst renamed to doc/flexible_efp.rst

File renamed without changes.

0 commit comments

Comments
 (0)