Skip to content

JDCurry/TIMP2-Docking-Pipeline

Repository files navigation

TIMP2 Docking and ADMET Pipeline

A computational framework for identifying novel allosteric modulators of TIMP2 and exploring mechanisms of neuroplasticity.


Preprint DOI

Overview

Please cite this work if you use or adapt the pipeline.

TIMP2 (Tissue Inhibitor of Metalloproteinase-2) regulates hippocampal plasticity and cognitive function through MMP-independent neuronal signaling.
Building on Ferreira et al. (2023, eNeuro) and integrin-mechanistic studies by Pérez-Martínez & Jaworski (2005) and Herrera-Molina et al. (2012), this project identifies a cryptic, chemically promiscuous allosteric pocket in TIMP2 and demonstrates that it accommodates both halogenated and non-halogenated scaffolds with high CNS drug-likeness.


Biological Context

TIMP2 modulates neural plasticity through dual integrin-dependent signaling systems that together maintain a dynamic balance between neurite growth and retraction.

  • Pérez-Martínez & Jaworski (2005) showed that TIMP2 activates neuronal α3β1 integrin, stimulating the cAMP/Rap1/ERK cascade and promoting neurite outgrowth and differentiation — a growth-promoting, pro-plasticity pathway independent of metalloproteinase inhibition.
  • Herrera-Molina et al. (2012) revealed that astrocytic αVβ3 integrin binds neuronal Thy-1, triggering Src inactivation and neurite retraction, defining an opposing, retractive mechanism relevant to axonal remodeling and glial responses.
  • Ferreira et al. (2023) confirmed in vivo that neuronal TIMP2 is essential for hippocampal neurogenesis and memory formation, directly linking these integrin pathways to cognitive function.

Together, these findings describe a bidirectional integrin–matrix feedback loop:
TIMP2–α3β1 drives neuronal growth and synaptic plasticity, whereas Thy-1–αVβ3 constrains or reverses it during glial activation.
The cryptic allosteric site discovered in this project may stabilize TIMP2’s pro-plasticity conformation and tip this balance toward regeneration.


Proposed Mechanism of Action

Proposed mechanism of action

Proposed mechanism of action for TIMP2 allosteric modulation.
Binding of a small molecule to the cryptic N-terminal pocket (Val6, Leu100, Phe103) induces a conformational shift enhancing TIMP2’s affinity for α3β1 integrins.
This promotes downstream activation of the cAMP/Rap1/ERK pathway, leading to increased neurite outgrowth, neurogenesis, and synaptic plasticity.

References:

  • Pérez-Martínez L. & Jaworski D.M. (2005). TIMP-2 promotes neuronal differentiation and neurite outgrowth through α3β1 integrin and ERK signaling. J Neurosci. 25 (26): 7997–8007.
  • Herrera-Molina R. et al. (2012). The integrin αVβ3 interacts with Thy-1 and regulates neurite outgrowth in hippocampal neurons. PLoS ONE 7 (4): e34295.
  • Ferreira T.A. et al. (2023). Neuronal TIMP2 regulates hippocampal neurogenesis and synaptic plasticity. eNeuro 10 (4): ENEURO.0031-23.2023.

Pipeline Summary

Step Script Function
1️⃣ dock_adaptive.py Adaptive AutoDock Vina screening with CPU benchmarking.
2️⃣ rebuild_hits.py Parses .pdbqt files to rebuild SMILES and docking scores.
3️⃣ timp2_triage.py Triage CNS/peripheral hits; energy and diversity filters.
4️⃣ timp2_analysis.py Integrates ADMET data, CNS MPO, and scaffold clustering.
5️⃣ add_zinc_ids_prioritized.py Re-maps prioritized compounds to ZINC IDs.
6️⃣ tier_1_filter.py Strict ADMET-AI toxicity tiers (Tier 1 = clean).
7️⃣ diversity_selection_script.py Balances chemical diversity across chunks.
8️⃣ singleton.py Identifies unique (singleton) scaffolds.
9️⃣ scaffold_extraction_rdkit.py Generates Murcko scaffold images.

Computational Setup

Hardware: Dual Xeon (24 cores, 128 GB RAM), NVMe SSD, Windows 10 Pro
Software: AutoDock Vina 1.2.5, RDKit 2023.03, OpenBabel 3.1.1, ADMET-AI v2.1, Schrödinger SiteMap (2025-2), PyMOL 2.6-3.1

Create environment:

conda env create -f environment.yml
conda activate timp2-docking

Example run:

python scripts/dock_adaptive.py data/strict_0001.smi
python scripts/rebuild_hits.py
python scripts/timp2_triage.py --hits-csv work_strict/strict_0001_hits.csv --out triage_out/strict_0001
python scripts/timp2_analysis.py triage_out/strict_0001/cns_hits.csv -o analysis_out/

About

CNS-targeted virtual screening of TIMP2 modulators with ADMET and toxicity profiling.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published