lacuna-pockets 0.3.0

Cryptic binding pocket discovery via conformational ensemble analysis

Introduction

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 a continuous crypticity score.

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 diffusion sampling (experimental, GPU)
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

# Optional Boltz-2 backend (experimental - see the Backends note)
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, default), OpenMM implicit-solvent MD, or experimental Boltz-2 diffusion sampling
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 crypticity by default; --rank-by druggability is available for always-open / orthosteric sites

---

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]	experimental	~100s/protein (GPU)	Boltz-2 diffusion sampling from sequence; high diversity but noisy (see note)
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 is the zero-dependency default.

Boltz backend status (honest note). The boltz backend runs Boltz-2 diffusion sampling on a GPU, but it currently predicts each conformer de novo from sequence (not partial diffusion from the input structure), which yields structurally divergent, noisy ensembles (150-300+ pocket clusters vs NMA's ~35). In GPU benchmarking it did not reliably improve cryptic detection over NMA. A proper apo-templated integration with sequence-based residue mapping is planned; until then, NMA is the recommended backend.

---

Benchmarks

7 / 22 cryptic pockets localized (32%, size-robust criterion; NMA backend, crypticity ranking, 20 conformers).

This curated result is cross-validated on two further independent datasets - PocketMiner 31% and CryptoBench 18% (the largest and hardest) - see Independent validation below.

Size-robust success criterion (top-5 pockets): a pocket whose lining residues reach a Jaccard overlap ≥ 0.25 with the known ligand-contact site (Jaccard = |found ∩ known| / |found ∪ known|), or whose center is within 4 Å of the site centroid. Lining residues use a true atomic-contact definition (any residue with an atom within 5 Å of the detected cavity). Reproduce with python benchmarks/cryptic_benchmark.py --category cryptic.

Why the number is lower than you may have seen before - please read. Earlier releases reported this benchmark using plain recall (|found ∩ known| / |known| ≥ 30%), which gave 13/22 (59%). That metric is size-gameable: a large pocket engulfs most of a small known site and scores high recall while sitting nowhere near it. We verified this directly - a learned re-ranker reached 84% on the recall metric purely by ranking pockets on raw volume. We therefore switched the headline to a size-robust criterion (Jaccard, which penalizes oversized pockets, OR a ≤4 Å centroid hit). Under it the honest numbers roughly halve. Both criteria are printed side by side by every benchmark script; we lead with the robust one because it is the number we can defend on held-out data.

Of the 22 cryptic targets, 2 pass the strict ≤4 Å centroid test (IL-2, PTP1B) and 5 more clear Jaccard ≥ 0.25. Precise pocket-center localization is genuinely hard for elongated, partially-open cryptic grooves, which is why the centroid-only pass rate is low. cryptic_benchmark.py prints the full per-metric breakdown (centroid, Jaccard at 0.20/0.25/0.30, and legacy recall).

Cryptic pockets - 7 / 22 (32%)

Sorted by Jaccard (size-robust overlap). ✅ = passes the size-robust criterion (Jaccard ≥ 0.25 or centroid ≤ 4 Å); recall is the legacy size-gameable metric, shown for contrast. "Rank" is the position of the best-matching top-5 pocket.

Protein	Apo PDB	Drug target	Jaccard	Recall	Rank
✅ BCL-XL BH3 groove	1LXL	navitoclax	56%	68%	1
✅ BCL-2 BH3 groove	1G5M	venetoclax	48%	59%	1
✅ MDM2 p53-binding cleft	1Z1M	nutlin-3	39%	47%	1
✅ PTP1B allosteric helix site	1A5Y	benzofurans	36%	94%	5
✅ IL-2 helix-α1 site	1M47	-	36%	93%	1
✅ HIV-1 RT NNRTI pocket	1HMV	nevirapine	33%	62%	4
✅ K-Ras switch-II pocket	4OBE	sotorasib/adagrasib	26%	79%	3
❌ Ricin A pterin pocket	1RTC	-	18%	50%	-
❌ T4 Lysozyme L99A cavity	1L90	-	17%	62%	-
❌ HCV NS5B thumb-site I	1NB4	VXR class	16%	47%	-
❌ Glucokinase allosteric site	1V4S	activators	15%	39%	-
❌ Src myristate pocket	2SRC	-	14%	36%	-
❌ PPARγ allosteric site	2PRG	metaglidasen	11%	35%	-
❌ c-ABL myristate pocket	3CS9	asciminib	7%	19%	-
❌ p38α DFG-out pocket	1P38	BIRB 796	7%	24%	-
❌ ERK2 allosteric site	2ERK	-	6%	19%	-
❌ Caspase-1 dimer interface	2HBQ	-	5%	25%	-
❌ PKM2 subunit interface	1ZJH	TEPP-46	4%	17%	-
❌ MMP-13 S1′ tunnel	2OZR	non-zinc	4%	6%	-
❌ TEM-1 allosteric site	1JWP	CBT	2%	17%	-
❌ IDH1 R132H dimer interface	3MAP	ivosidenib	2%	7%	-
❌ SHP-2 allosteric tunnel	2SHP	SHP099	0%	0%	-

The remaining gap is mostly sampling, not ranking. Raising the cutoff from top-5 to top-20 lifts the size-robust score only from 7/22 to 10/22 - just 3 pockets are detected-but-mis-ranked. The other 12 misses are not localized at all even at top-20, so they are a sampling/localization ceiling (the NMA ensemble never opens or the detector never localizes the site tightly enough) rather than a ranking failure. This is the honest picture: under the older recall metric the top-20 ceiling looked like 73%, which made the problem appear to be ranking - it was largely the metric. The hard cases split into oligomeric-interface pockets (Caspase-1, IDH1, PKM2) that form between subunits and are invisible to single-chain analysis, and large-rearrangement sites (p38 DFG-out, c-ABL myristate) that need sampling beyond elastic-network modes.

Dimer-interface pockets are partly addressable with --homodimer (reads BIOMT records and builds the biological assembly), though this benchmark's single-chain-referenced scoring does not credit them. For large-rearrangement sites the optional Boltz-2 backend samples more broadly, but its current sequence-based integration is noisy - see Backends.

Independent validation - three benchmarks

Measured on three independent datasets (NMA + crypticity, top-5). Both criteria are reported: the size-robust headline (Jaccard ≥ 0.25 or ≤ 4 Å centroid) and the legacy recall number (≥ 30% recall or ≤ 4 Å centroid) that earlier releases led with.

Benchmark	N	Size-robust	Legacy recall	Notes
Curated apo/holo set (this repo)	22	32%	59%	literature cryptic pairs
PocketMiner (Meller 2023, Nat. Commun.)	45	31%	60%	per-residue cryptic labels
CryptoBench test fold (Vavra 2024, Bioinformatics)	180	18%	49%	largest & most diverse; harder

The two curated/field-standard sets converge at ~31-32% under the size-robust metric; CryptoBench - the field's largest cryptic set (1107 structures; 180 of its 222-structure held-out test fold evaluated here) - is harder at 18%. The legacy recall column roughly doubles every number: that gap is the size-gaming headroom the recall metric leaves open (a large pocket covers a small known site without being localized on it), which is exactly why the size-robust number is the one we lead with. Reproduce (each script prints both criteria):

python benchmarks/pocketminer_benchmark.py    # PocketMiner (auto-downloads)
python benchmarks/cryptobench_benchmark.py    # CryptoBench test fold (auto-downloads, ~10 min)

Orthosteric / conformational controls

Crypticity ranking (the default) intentionally de-prioritizes always-open sites, so for orthosteric / general pocket finding use --rank-by druggability. Under the corrected contact-lining pipeline (NMA, --rank-by druggability):

Category	Result	Notes
Orthosteric	3 / 6	hen lysozyme 100%, HIF-2α 96% (1.1 Å centroid), DHFR 50%; misses HIV protease, thrombin, trypsin (1S0Q numbering)
Conformational	1 / 1	adenylate kinase open→closed

Orthosteric detection is a known relative weakness of the tight-contact pipeline - the tool is tuned for transient cryptic sites, not always-open active-site grooves.

Crypticity score

Every reported pocket 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. Ranking by crypticity is the default and recovers the most cryptic targets. 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
crypticity (default)	most cryptic sites first	12 / 20
druggability	peak open-state composite druggability	10 / 20
balanced	druggability with a mild persistence bonus	8 / 20
persistence	legacy persistence × druggability	7 / 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	✅ 79%, 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 --category cryptic   # 22 cryptic targets, NMA (~4 min)
python benchmarks/cryptic_benchmark.py --quick              # 10 conformers, faster
python benchmarks/cryptic_benchmark.py --category cryptic --rank-by druggability  # ablation
python benchmarks/cryptic_benchmark.py --category cryptic --top-n 20              # detection ceiling
python benchmarks/compare_fpocket.py                        # fpocket head-to-head

---

Example: K-Ras switch-II

# Download K-Ras apo (from RCSB); NMA backend (default) recovers switch-II at rank 3
lacuna discover 4OBE.pdb \
    --conformers 20 \
    --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.