Biologically-Inspired Examples

This document showcases how to use synth-pdb to generate structures that mimic real biological proteins and motifs.

1. Human Epidermal Growth Factor (EGF)

EGF is a small protein (53 residues) with a complex disulfide-bonded structure. It contains three critical disulfide bonds (C6-C20, C14-C31, C33-C42) that are essential for its biological activity.

You can generate an EGF-like structure using the following command:

python -m synth_pdb.main \
  --sequence "NSDSECPLSHDGYCLHDGVCMYIEALDKYACNCVVGYIGERCQYRDLKWWELR" \
  --conformation random \
  --minimize \
  --cap-termini \
  --gen-shifts \
  --gen-relax \
  --output egf_protein.pdb

Key Features of this Example: - Sequence: Uses the actual human EGF sequence. - Minimization: The --minimize flag triggers OpenMM to relax the structure and automatically detect/model the disulfide bonds. - NMR Data: Generates synthetic chemical shifts and relaxation data (--gen-shifts, --gen-relax). - Biophysical Realism: Adds N- and C-terminal caps (--cap-termini).

2. Zinc Finger Motif

Zinc fingers are common structural motifs in proteins that coordinate one or more zinc ions to help stabilize their fold.

synth-pdb can automatically detect Zinc-binding motifs (like CCHH or CCCC) and insert a Zinc ion:

python -m synth_pdb.main \
  --sequence "PYKCPECGKSFSQKSDLVKHQRTHTG" \
  --conformation random \
  --minimize \
  --metal-ions auto \
  --output zinc_finger.pdb

The --metal-ions auto flag scans the sequence for known binding motifs and places the ZN ion in the geometric center of the coordinating residues.

3. Cyclic Peptides

Many natural antibiotics and hormones are cyclic peptides. synth-pdb can generate head-to-tail cyclic structures:

python -m synth_pdb.main \
  --sequence "C-G-G-C" \
  --cyclic \
  --minimize \
  --output cyclic_peptide.pdb

The --cyclic flag ensures that the N-terminal Nitrogen and C-terminal Carbon are bonded, and the minimizer relaxes the structure into a stable cyclic conformation.

4. D-Amino Acid Peptides

D-amino acids are often found in bacterial cell walls and specialized peptides. They are "mirror images" of the standard L-amino acids and can impart resistance to proteases.

python -m synth_pdb.main \
  --sequence "DAL-GLY-GLY-DPH" \
  --output d_peptide.pdb

synth-pdb supports the standard PDB 3-letter codes for D-amino acids (e.g., DAL for D-Alanine, DPH for D-Phenylalanine).