lacuna-pockets 0.2.0

Cryptic binding pocket discovery via conformational ensemble analysis

Lacuna

Cryptic binding pocket discovery via conformational ensemble analysis.

Most protein structure predictors (AlphaFold, Boltz, Chai) give you one static structure. But ~70% of disease-relevant proteins are considered "undruggable" not because they're biologically intractable - it's because no pocket is visible in their ground state. K-Ras was "undruggable" for 30 years until a transient cryptic pocket was found in its switch-II region. That pocket now backs sotorasib and adagrasib.

Lacuna finds those pockets. It generates a conformational ensemble from any input structure, detects pockets per conformer, and clusters them across the ensemble to surface sites that only appear transiently - ranked by druggability and persistence.

lacuna discover kras.pdb --conformers 20 --emit-boltz-constraints --emit-vina-boxes

---

Install

pip install lacuna-pockets

Optional backends (better conformational sampling):

pip install "lacuna-pockets[openmm]"   # 100ps implicit-solvent MD
pip install "lacuna-pockets[boltz]"    # Boltz-2 partial diffusion (best quality, GPU recommended)
pip install "lacuna-pockets[all]"      # everything

Requires Python 3.10+.

---

Quick start

CLI

# Discover pockets with defaults (NMA backend - physically grounded, no GPU needed)
lacuna discover protein.pdb --conformers 20

# Filter and limit output
lacuna discover protein.pdb --min-druggability 0.5 --min-persistence 0.3 --top 5

# Analyze a homodimer - detects pockets at the dimer interface (e.g. Caspase-1, IDH1)
# Reads BIOMT records from PDB; for best results use the biological assembly download from RCSB
lacuna discover protein.pdb --homodimer --conformers 20

# Use Boltz-2 partial diffusion for highest-quality sampling
lacuna discover protein.pdb --backend boltz --conformers 30

# Emit all docking file formats
lacuna discover protein.pdb --emit-boltz-constraints --emit-vina-boxes --emit-pocket-pdbs

# Generate docking files from a previous report
lacuna dock-prep kras_lacuna/pocket_report.json kras.pdb --format all

Python API

from lacuna import load_structure, detect_pockets, cluster_pockets
from lacuna.ensemble.nma_backend import NMABackend
from lacuna.io.structure import coords_array
from lacuna.io.writers import write_report, write_boltz_constraint

structure = load_structure("protein.pdb")
backend = NMABackend(seed=42)
coord_sets = backend.generate("protein.pdb", n_conformers=20)

all_coords = [coords_array(structure)] + coord_sets
pocket_lists = []
for ci, coords in enumerate(all_coords):
    pockets = detect_pockets(coords, structure)
    for p in pockets:
        p.conformer_idx = ci
    pocket_lists.append(pockets)

clusters = cluster_pockets(pocket_lists, n_conformers=len(all_coords))
for c in clusters[:5]:
    print(f"Rank {c.rank}  druggability={c.druggability:.3f}  "
          f"persistence={c.persistence:.0%}  cryptic={c.cryptic}")
    print(f"  Residues: {', '.join(c.lining_residues[:5])}")

---

How it works

1. Ensemble generation - Generate N conformers via elastic network model normal mode analysis (built-in), OpenMM implicit-solvent MD, or Boltz-2 partial diffusion at varying noise levels
2. Pocket detection - Grid-based alpha-point analysis per conformer: compute distance transform, find local maxima within the 1.4–5.5 Å interaction zone, cluster nearby alpha-points into pocket candidates
3. Cross-ensemble clustering - Greedy centroid merging clusters corresponding pockets across all conformers
4. Druggability scoring - Gaussian volume reward centered at 300 ų + enclosure + hydrophobicity + aromaticity (Halgren 2009), scored in each conformer
5. Crypticity scoring & ranking - Each site gets a continuous crypticity score (how much it opens relative to the apo state × druggability when open) and is flagged cryptic: true if present in <90% of conformers. Pockets are ranked by peak open-state druggability by default; --rank-by crypticity surfaces the most cryptic sites first

---

Outputs

File	Description
pocket_report.json	Ranked pocket metadata: centroid, volume + apo→open range, druggability, crypticity, persistence, lining residues
pocket_N_site.pdb	Pseudoatom PDB for PyMOL/ChimeraX visualization
pocket_N_constraint.yaml	Boltz YAML - add a SMILES and run boltz predict to dock into this site
pocket_N_vina.conf	AutoDock Vina / Gnina / QuickVina box config

---

Backends

Backend	Install	Quality	Speed	Notes
nma	built-in	good	~0.1s/conf	Anisotropic Network Model normal mode analysis - hinge bending, breathing, twist motions
openmm	lacuna[openmm]	good	~2s/conf	100ps Langevin MD, GBn2 implicit solvent
boltz	lacuna[boltz]	best	~30s/conf (GPU)	Boltz-2 partial diffusion at varying noise fractions
random	built-in	baseline	~0.04s/conf	Correlated Gaussian backbone perturbation

Auto-selection order: boltz → openmm → nma → random. On a plain pip install lacuna, the NMA backend runs automatically.

The nma backend samples physically meaningful collective motions — the same hinge-bending and breathing modes that open cryptic pockets in nature — without requiring a GPU or force field. It replaces random as the zero-dependency default. For large-scale loop rearrangements or very deep cryptic sites, boltz remains the best option.

---

Benchmarks

17 / 20 cryptic pockets detected (85%, NMA backend, 20 conformers) - exceeding the CryptoSite published benchmark rate on a statistically defensible N=20 set.

Success criterion (field standard, top-5 pockets): pocket centroid within 4 Å of the known binding-site centroid or ≥30% residue overlap. The numbers below are reproduced by the default configuration (--backend nma --rank-by druggability).

Transparency: these are the OR-criterion pass counts. Reported per-metric: of the 20 cryptic targets, 17 pass on residue overlap and 2 also satisfy the stricter centroid-distance test. The centroid-of-binding-residues is an intentionally strict and somewhat ill-posed reference for elongated grooves, so the residue-overlap criterion (used by CryptoSite and PocketMiner) is the primary metric. cryptic_benchmark.py now prints the full per-metric breakdown.

Cryptic pockets - 17 / 20 (85%)

Protein	Apo PDB	Drug target	Overlap	Rank	Time
✅ T4L L99A hydrophobic cavity	1L90	-	100%	1	0.9s
✅ K-Ras switch-II pocket	4OBE	sotorasib / adagrasib	93%	3	0.9s
✅ IL-2 helix-α1 site	1M47	-	100%	4	0.7s
✅ MDM2 p53-binding cleft	1Z1M	nutlin-3	47%	4	1.1s
✅ BCL-XL BH3 groove	1LXL	navitoclax	91%	2	2.9s
✅ BCL-2 BH3 groove	1G5M	venetoclax	96%	5	1.2s
✅ c-ABL myristate pocket	3CS9	asciminib	44%	1	1.7s
✅ PTP1B allosteric helix site	1A5Y	benzofuran inhibitors	59%	3	2.2s
✅ p38α DFG-out pocket	1P38	BIRB 796	38%	1	3.0s
✅ HIV-1 RT NNRTI pocket	1HMV	nevirapine	94%	2	8.4s
✅ HCV NS5B thumb-site I	1NB4	VXR class	60%	5	4.7s
✅ PPARγ allosteric AF-2 site	2PRG	metaglidasen	65%	4	1.7s
✅ Glucokinase allosteric site	1V4S	B84 activator	100%	3	3.8s
✅ MMP-13 S1′ allosteric tunnel	2OZR	non-zinc inhibitors	56%	3	1.1s
✅ Src myristate pocket	2SRC	-	48%	5	4.0s
✅ SHP-2 allosteric tunnel	2SHP	SHP099 class	47%	1	6.1s
✅ ERK2 allosteric site	2ERK	-	31%	1	2.9s
❌ Caspase-1 dimer interface	2HBQ	-	8%	-	1.7s
❌ IDH1 R132H dimer interface	3MAP	ivosidenib	21%	-	3.7s
❌ PKM2 subunit-interface activator	1ZJH	TEPP-46 class	25%	-	4.3s

Switching from the zero-dependency random backend to the physics-based nma backend recovers the former near-misses (IL-2 21%→100%, Src 28%→48%, SHP-2 and ERK2 now pass), and ranking by peak open-state druggability surfaces the right pocket into the top 5.

All three remaining misses are oligomeric-interface pockets - they form between subunits and cannot be seen in a single-chain analysis. They are addressable with the --homodimer flag, which reads BIOMT symmetry records and constructs the full biological assembly before analysis:

lacuna discover 2HBQ.pdb --homodimer --conformers 20   # Caspase-1 dimer interface
lacuna discover 3MAP.pdb --homodimer --conformers 20   # IDH1 R132H dimer interface

For the very hardest sites requiring large loop rearrangement, the optional Boltz-2 backend (--backend boltz) samples states unreachable by NMA.

Conformational and orthosteric controls

Category	Result	Notable entries
Conformational	1 / 1 (100%)	Adenylate kinase open→closed (rank 1)
Orthosteric	5 / 6 (83%)	HIF-2α 100% (1.1 Å centroid), lysozyme 100%, thrombin, DHFR 72%
Orthosteric miss	-	Trypsin (1S0Q non-standard residue numbering - documented limitation)

Overall across all 27 proteins: 23 / 27 (85%).

Crypticity score

Every reported pocket now carries a continuous crypticity score in [0, 1] - the conformational-selection signature of a cryptic site, defined as how much the pocket opens relative to the apo/input structure × how druggable it is once open:

opening    = (max_volume − apo_volume) / max_volume        # 1.0 if absent in the apo state
crypticity = opening × peak_open_state_druggability

A constitutive pocket already formed in the input structure scores ≈ 0; a pocket absent in the apo structure that opens into a druggable cavity scores near 1. As an independent validation, ranking the benchmark purely by crypticity (--rank-by crypticity) still recovers 15/20 known cryptic pockets - the score discriminates true cryptic sites with no druggability tie-breaking. The JSON report also includes per-pocket volume dynamics (apo_volume_A3, volume_range_A3) and max_druggability.

Ranking strategies

--rank-by selects how pockets are ordered (cryptic benchmark pass rate, NMA, N=20):

Strategy	Description	Cryptic pass
druggability (default)	peak open-state composite druggability	17 / 20
persistence	legacy persistence × druggability	16 / 20
balanced	druggability with a mild persistence bonus	15 / 20
crypticity	most cryptic sites first	15 / 20

Speed (NMA backend, no GPU)

Protein size	Time
~130 residues (lysozyme)	0.6s
~170 residues (MDM2)	1.1s
~350 residues (K-Ras)	0.9s
~530 residues (HIV-1 RT chain A)	8.4s

Head-to-head: Lacuna vs fpocket

fpocket detects pockets on a single static structure. Lacuna generates a conformational ensemble - the critical difference for cryptic sites that are absent in the apo crystal.

Target	fpocket 4.2	Lacuna (NMA backend)
1HEL hen lysozyme (orthosteric)	✅ rank 1	✅ 100%
1L90 T4L L99A (cryptic)	❌ not in top 5	✅ 100%, rank 1
4OBE K-Ras switch-II (cryptic)	❌ not in top 5	✅ 93%, rank 3
1HPV HIV-1 protease (orthosteric)	✅ rank 1	✅ rank 1
Score	2 / 4	4 / 4

T4L L99A and K-Ras switch-II are the canonical single-structure benchmark failures: the T4L cavity is physically absent in the apo crystal (<100 ų), and the K-Ras switch-II pocket only opens during nucleotide exchange.

Reproduce:

python benchmarks/cryptic_benchmark.py          # full 27-protein run, NMA backend (~5 min)
python benchmarks/cryptic_benchmark.py --quick  # 10 conformers (~2 min)
python benchmarks/cryptic_benchmark.py --category cryptic            # cryptic only
python benchmarks/cryptic_benchmark.py --backend random --rank-by persistence  # ablations
python benchmarks/compare_fpocket.py            # fpocket head-to-head

---

Example: K-Ras switch-II

# Download K-Ras apo (from RCSB)
# Run with Boltz backend for highest-quality switch-II sampling
lacuna discover 4OBE.pdb \
    --backend boltz \
    --conformers 30 \
    --emit-boltz-constraints \
    --output kras_pockets/

# pocket_0_constraint.yaml is ready - add your SMILES:
#   - ligand:
#       id: L
#       smiles: YOUR_SMILES_HERE
boltz predict kras_pockets/pocket_0_constraint.yaml

See examples/kras_cryptic.py for a full annotated Python workflow.

---

Input formats

Accepts PDB or mmCIF from any predictor or database:

• AlphaFold 2 / AlphaFold 3
• Boltz-1 / Boltz-2
• Chai-1
• RCSB PDB
• ESMFold, RoseTTAFold, OpenFold, etc.

---

Citation

If you use Lacuna in published research, please cite:

Moore, C. (2026). Lacuna: Cryptic Binding Pocket Discovery via Conformational Ensemble Analysis. https://github.com/mooreneural/lacuna

BibTeX:

@software{moore2026lacuna,
  author  = {Moore, Clayton W.},
  title   = {Lacuna: Cryptic Binding Pocket Discovery
             via Conformational Ensemble Analysis},
  year    = {2026},
  url     = {https://github.com/mooreneural/lacuna},
  version = {0.2.0}
}

Methodology papers Lacuna builds on:

• Atilgan et al. (2001) Biophys. J. 80(1):505–515 - Anisotropic Network Model (NMA backend)
• Halgren (2009) J. Chem. Inf. Model. 49(2):377–389 - SiteMap druggability scoring
• Le Guilloux et al. (2009) BMC Bioinformatics 10:168 - fpocket alpha-sphere approach
• Schmidtke & Barril (2010) J. Med. Chem. 53(15):5858–5867 - enclosure scoring

---

License

GNU AGPL-3.0-or-later — free to use, study, modify, and share. The AGPL's copyleft requires that if you distribute a modified version, or run a modified version as a network/hosted service, you make the complete corresponding source available under the same license.

A separate commercial license removes the AGPL copyleft obligation (for embedding Lacuna in closed-source products or hosted services without releasing your own source) and adds warranty, indemnification, support SLAs, and custom development. Contact [e-mail].

Versions ≤ 0.1.2 were released under the MIT License and remain available under those terms. AGPL-3.0 applies from version 0.2.0 onward.