primacito/src/legacy/export.py

206 lines
7.2 KiB
Python

import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from backbone import generate_backbone_chain
from nerf import place_atom_nerf
from ramachandran import RamachandranSampler
from sidechain import SIDECHAIN_TOPOLOGY, build_full_atom_chain, sample_chi_angle
def add_backbone_oxygens(chain, nerf_func):
"""
Adds the Carbonyl Oxygen (O) to every residue.
The C=O bond bisects the N-C-CA angle or is planar to the peptide bond.
"""
for i, res in enumerate(chain):
# We place O relative to C, CA, and N.
# Ref: C (parent), CA (grand), N (great-grand)
# Bond: 1.23 A (Double bond)
# Angle: ~121 degrees
# Torsion: 180 (Planar, trans to N-CA bond)
o_coord = nerf_func(
res['C'], res['CA'], res['N'],
bond_length=1.23,
bond_angle_deg=121.0,
torsion_angle_deg=180.0
)
# We store it in sidechain list for convenience in plotting/writing
if 'sidechain' not in res:
res['sidechain'] = []
# Insert at the beginning of sidechain list so it appears after C in PDB
res['sidechain'].insert(0, {'name': 'O', 'coord': o_coord})
return chain
def format_atom_name(name):
"""
PDB atom names must be 4 chars.
- 4-char names (HD21) occupy cols 13-16.
- <4 char names (CA, N, O) start at col 14.
"""
if len(name) == 4:
return name
else:
return f" {name:<3}" # " CA "
def save_to_pdb(chain, filename="unfolded_structure.pdb"):
"""
Writes the chain to a standard PDB format file.
"""
atom_serial = 1
with open(filename, 'w') as f:
f.write("REMARK 1 GENERATED BY UNFOLDED POLYPEPTIDE GENERATOR\n")
for i, res in enumerate(chain):
res_name = res['name']
chain_id = 'A'
res_seq = i + 1
# 1. Write Backbone Atoms in Order: N, CA, C
bb_order = ['N', 'CA', 'C']
# Combine backbone and sidechain into one list for writing
# (Backbone first, then sidechain)
all_atoms = []
for atom_key in bb_order:
all_atoms.append({'name': atom_key, 'coord': res[atom_key]})
# Add sidechain atoms (including O, H, etc.)
# Note: O is already inserted into 'sidechain' by our helper
if 'sidechain' in res:
all_atoms.extend(res['sidechain'])
for atom in all_atoms:
name = atom['name']
x, y, z = atom['coord']
# Determine Element (First letter of name usually)
element = name[0]
# PDB fixed-width format
# ATOM (1-6)
# Serial (7-11)
# Name (13-16)
# ResName (18-20)
# Chain (22)
# SeqNum (23-26)
# X, Y, Z (31-54)
# Occ (55-60), Temp (61-66)
# Element (77-78)
pdb_line = (
f"ATOM {atom_serial:>5} {format_atom_name(name):<4} {res_name:>3} "
f"{chain_id:>1}{res_seq:>4} "
f"{x:>8.3f}{y:>8.3f}{z:>8.3f}"
f"{1.00:>6.2f}{0.00:>6.2f} {element:>2}\n"
)
f.write(pdb_line)
atom_serial += 1
f.write("END\n")
print(f"Successfully saved PDB to: {filename}")
def visualize_structure(chain, mode='backbone'):
"""
Visualizes the protein structure.
Modes: 'ca_trace', 'backbone', 'all_atom'
"""
fig = plt.figure(figsize=(12, 10))
ax = fig.add_subplot(111, projection='3d')
# Extract CA coords for the main trace
ca_x = [res['CA'][0] for res in chain]
ca_y = [res['CA'][1] for res in chain]
ca_z = [res['CA'][2] for res in chain]
# Plot the main chain line
ax.plot(ca_x, ca_y, ca_z, color='grey', linewidth=2, label='Backbone Trace')
if mode == 'ca_trace':
ax.scatter(ca_x, ca_y, ca_z, s=50, c='blue', label='CA')
elif mode == 'backbone':
# Show N, CA, C
for res in chain:
ax.scatter(*res['N'], c='blue', s=30)
ax.scatter(*res['CA'], c='green', s=30)
ax.scatter(*res['C'], c='red', s=30)
elif mode == 'all_atom':
# Plot everything including sidechains
for res in chain:
# Backbone
ax.scatter(*res['N'], c='blue', s=20, alpha=0.6)
ax.scatter(*res['CA'], c='green', s=20, alpha=0.6)
ax.scatter(*res['C'], c='red', s=20, alpha=0.6)
if 'sidechain' in res:
for atom in res['sidechain']:
coord = atom['coord']
name = atom['name']
color = 'orange'
size = 10
if name.startswith('H'):
color = 'lightgrey'
size = 5
elif name.startswith('O'):
color = 'red'
size = 20
elif name.startswith('N'):
color = 'blue'
size = 20
elif name.startswith('S'):
color = 'yellow'
size = 30
ax.scatter(coord[0], coord[1], coord[2], c=color, s=size)
# Aesthetics
ax.set_xlabel('X ($\AA$)')
ax.set_ylabel('Y ($\AA$)')
ax.set_zlabel('Z ($\AA$)')
ax.set_title(f'Generated Structure ({len(chain)} residues) - Mode: {mode}')
plt.legend()
plt.show()
if __name__ == "__main__":
# --- 1. INITIALIZATION ---
# Import your previous classes/functions here
# from generator import generate_backbone_chain, build_full_atom_chain, ...
# For now, assuming they are in the same memory space
sampler = RamachandranSampler()
# --- 2. DEFINE SEQUENCE ---
# A nice peptide with diverse chemistry for electrostatic testing
# (Charged, Polar, Hydrophobic, Glycine, Proline)
sequence = ['MET', 'LYS', 'ASP', 'GLY', 'PHE', 'ARG', 'GLU', 'VAL', 'HIS', 'ALA']# * 1000
print(f"Generating unfolded structure for: {'-'.join(sequence)}")
# --- 3. GENERATE BACKBONE ---
# (Generates N, CA, C)
backbone, total_backtracks = generate_backbone_chain(sequence, place_atom_nerf, sampler, verbose=True)
# --- 4. BUILD SIDECHAINS & PROTONS ---
# (Generates Sidechains, HA, HB.., NH3+ term, COO- term)
full_struct = build_full_atom_chain(backbone, SIDECHAIN_TOPOLOGY, place_atom_nerf, sample_chi_angle)
# --- 5. ADD CARBONYL OXYGENS ---
# (Generates O on the backbone C)
full_struct = add_backbone_oxygens(full_struct, place_atom_nerf)
# --- 6. EXPORT TO PDB ---
save_to_pdb(full_struct, filename="unfolded_protein.pdb")
# --- 7. VISUALIZE ---
# 'all_atom' shows the hydrogens and sidechains
#visualize_structure(full_struct, mode='all_atom')