Keywords: Q-MOL, protein flexibility treatment, protein-ligand docking, intrinsically disordered proteins, IDP, protein folding funnel, energy landscape theory, ELT, prediction of allosteric sites, prediction of cryptic binding pockets, no well-defined pockets needed, flexible proteins, undruggable protein targets

What is Q-MOL?

Q-MOL is the universal in silico small molecule drug discovery platform for de novo discovery of hits against all types of protein targets. 

See preprint Q-MOL: High Fidelity Platform for In Silico Drug Discovery and Design for details.

Q-MOL is the middle ground between classical full atom modeling and a black box of a trained neuronal network.  The Q-MOL solution combines a full atom thermodynamically correct molecular mechanics simulation and machine learning compute components that augment full atom protein structure presentation.

The primary and the unique feature of the Q-MOL platform is the ability to correctly predict small molecule binders against highly flexible and intrinsically disordered proteins (IDPs).

Intrinsically disordered and highly flexible proteins constitute the bulk of important drug targets. These include protein families of transcription factors, protein adaptors, flexible regulatory domains and interfaces of large enzymes, and many others. Only a few of these protein targets with known structure contain canonical druggable binding pockets. The rest of the bulk are considered undruggable, though it is understood that many of them may contain "cryptic" binding sites that are realized in specific protein conformations.

Q-MOL was specifically designed to achieve ultimate success in targeting these types of proteins.

Unique Q-MOL features

• Efficient protein-ligand docking with full implicit protein flexibility treatment of target protein
• No requirement for well-defined binding pockets, Q-MOL can dock ligands to flat structureless interfaces
• Correct location prediction of allosteric and cryptic binding sites on 3D structure of target protein
• Efficient protein-ligand docking against highly-flexible and intrinsically disordered proteins (targeting “undruggable” proteins)
• Every single Q-MOL feature was validated in real biological experiments

How does Q-MOL accomplish this?

• Q-MOL implicitly incorporates full protein flexibility into a protein-ligand docking simulation
• Q-MOL correctly detects allosteric and cryptic binding sites on the molecular surface of non-enzyme targets
• Every single computational feature of Q-MOL was validated in real biological experiments where predicted ligands were tested in vitro, in cellulo and in vivo (when possible)

Methodology Validation in Real Drug Discovery Projects

Unlike other drug discovery programs that are validated virtually, every single computational feature of Q-MOL was validated experimentally during development.

The Q-MOL in silico drug discovery platform was validated on the following protein targets (partial list) with 100% success rate.  The allosteric type of selected binding site is indicated with allosteric, types of assays are indicated.

• Furin proteinase calcium-dependent serine endoprotease:    active site, biochemical
• Dengue viral protease DV2 NS3/NS3B:    allosteric, biochemical, cell-based replicons
• WNV NS3/NS2B West Nile virus protease:    allosteric, biochemical, cell-based replicons, published
• HCV NS3/NS4A Hepatitis C viral protease:    allosteric, biochemical, cell-based replicons, published
• Zika viral protease:   allosteric, biochemical, co-crystal structure, published
• LYP phosphatase:   active site, allosteric, biochemical
• LMWPTP Low Molecular Weight phosphatase:    active site, biochemical, cell-based
• RXR retinoid receptor α:   allosteric, biochemical, cell-based, published
• NADD bacterial NaMN adenylyltransferase:    active site, allosteric, biochemical
• GCYH-IB prokaryotic GTP cyclohydrolase:          active site, allosteric, bacterial cells, NIAD poster
• MT1-MMP PEX14 Matrix metalloproteinase 1 Hemopexin domain:    allosteric, biochemical, cell-based, animal model, published
• TNFR tumor necrosis factor receptor:    allosteric, biochemical, cell-based
• Rad51 DNA repair protein:   active site, allosteric, biochemical
• Rgs9 regulator of G-protein signaling:   allosteric, cell-based
• cMyc oncogene:   allosteric, biochemical, cell-based read the story COLLABORATORS WANTED!
• β-catenin beta-catenin Wnt signaling pathway:    allosteric, biochemical, cell-based, animal model, published
• β-arrestin1 protein regulation of G protein-coupled receptor (GPCR) signaling:    allosteric, biochemical
• Fbw7 F-box domain of ubiquitin ligase neuroprotective axis (Parkinson’s disease):    allosteric, cell-based, co-crystal structure, animal model
• Scp1 small C-terminal domain phosphatase 1:   allosteric, biochemical, cell-based
• Sox2 transcription factor embryonic development and maintaining stem cell pluripotency:    allosteric, cell-based

Obtaining Q-MOL

Contact Dr. Anton Cheltsov at anton at q-mol.com or chelton66 at proton.me.

Certain Q-MOL features are also available via functional Q-MOL portal https://q-mol.org

Anton Cheltsov PhD  anton at q-mol.com