Library · Simulation & modelling
What Structural Docking & Molecular Mechanics got wrong, from 37 theses
Summary written by a model from the records below: The records document a wide range of computational challenges encountered in structural docking and molecular mechanics protocols across diverse chemical and biological targets. Computational predictions frequently suffer from rigid-body approximations, scoring function inaccuracies compared to wet-lab measurements, and pose instability during simulation. These records come from graduate theses, 2021 to 2026, each labeled with its degree where the repository states it, and each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Plus 7 problems the authors left open, listed at the end.
Failure modes, by theses reporting them
Select a mode to jump to its records.
Advanced and learned docking tools are beaten by standard baseline methods
Established physics-based methods and simpler algorithms consistently outperformed newer neural network or geometric frameworks in docking benchmark evaluations. Tools such as Monte Carlo sampling, native ZDOCK, and FRODock achieved higher success rates and lower root-mean-square errors than specialized learned alternatives.
Lost to a baseline
Base RFAA achieved 42% docking success on PoseBusters, failing certain stringent atomic clash/distance checks compared to specialized geometry tools.
Explainable Machine Learning and Applications in Protein-Ligand Complex Structure Prediction · University of Washington · PhD
Lost to a baseline
AutoDock Vina CAMEO server baseline achieved only 8% docking success (< 2 Å RMSD) compared to 32% for RFAA.
Diffusion Models for Protein Structure Design: From Backbone Generation to Atomic-Resolution Enzyme Design · University of Washington · PhD
Lost to a baseline
Glide docking scoring function beat global Dragnet on Set 384 CDK2 (Glide r^2 0.62 vs global Dragnet r^2 0.12) and Set 1164 TNK2 (Glide r^2 0.59 vs global Dragnet r^2 0.86 with Glide MUE not listed, but MM-GBSA r^2 0.67 beats global Dragnet MUE 2.22 / r^2 0.86).
Building Blocks of Neural Network Intermolecular Interaction Potentials · Georgia Tech · PhD
Lost to a baseline
Monte Carlo docking generated stable complex structures for 113 pairs, whereas Voronoi docking generated stable structures for 107 pairs out of 120.
First-principles control of zeolite synthesis, transformations, and intergrowth · MIT · PhD
Lost to a baseline
RFAA docking accuracy was beaten significantly by AlphaFold 3 across ligand docking benchmarks.
Explainable Machine Learning and Applications in Protein-Ligand Complex Structure Prediction · University of Washington · PhD
Lost to a baseline
In oracle iRMSD on the rigid docking test set, FRODock achieved a median of 1.9 Å compared to DiffMaSIF's 3.5 Å
Predicting protein interactions using geometric deep learning on protein surfaces · EPFL · PhD
Lost to a baseline
On the holo/bound large-scale docking benchmark (100 targets), ZDock+ZRank2 solved 45 complexes in Top 1 (and 63 in Top 10) compared to MaSIF-search (2000 decoys) solving 43 in Top 1 (and 56 in Top 10)
Predicting protein interactions using geometric deep learning on protein surfaces · EPFL · PhD
Lost to a baseline
On unbound (apo) docking benchmark top-10, ZDock (5 solved) and ZDock+ZRank2 (5 solved) outperformed MaSIF-search (2 solved)
Predicting protein interactions using geometric deep learning on protein surfaces · EPFL · PhD
Tried and failed
unsupervised word2vec molecular embeddings as sole features applied to drug-target docking score prediction. Outcome: worse than baseline. Reason: mol2vec features alone lacked sufficient structural and chemical specificity compared to concatenated domain-specific fingerprints
Advancing Personalized Medicine Through Generative Artificial Intelligence · Georgia Tech · PhD
Lost to a baseline
DOVE (CNN docking tool) was beaten by native ZDOCK on reranking antibody model docking poses.
Deep learning algorithms for predicting association between antibody sequence, structure, and antibody properties · Oxford · PhD
Lost to a baseline
On Smina redocking, PoseTriager trained on Augmented2020 (MCC 0.34) was beaten by PoseTriager trained on standard Redocked2020 (MCC 0.44 ± 0.02).
Towards addressing structural data limitations in machine learning for small molecule drug discovery · Oxford · PhD
Rigid receptor docking fails due to steric clashes and lack of conformational flexibility
Treating receptors or bulky ligands as rigid structures caused severe steric exclusion and generated negative fitness scores across multiple docking programs. Static representations failed to accommodate induced-fit adaptations such as side-chain rotations, flexible peptide backbones, and transient pocket openings.
Considered and rejected
Considered and rejected: Single static docking snapshots / docking binding free energies rejected due to inability to capture protein induced fit, binding kinetics, and solvent entropy.
INTEGRATING ELECTRONIC STRUCTURE AND DYNAMIC STABILITY IN A PREDICTIVE MODEL FOR THE TOXICITY OF DIOXIN-LIKE COMPOUNDS · Saskatchewan · PhD
Considered and rejected
Considered and rejected: AutoDock Vina and HADDOCK docking were rejected for modeling the LptH-thanatin complex due to lack of backbone flexibility and generating unfolded/unrealistic peptide poses
Discovery and Characterization of Novel Thanatin Orthologs Against Escherichia coli LptA and Pseudomonas aeruginosa LptH · Duke · Doctorate
Considered and rejected
Considered and rejected: Rejected rigid-body Fourier transform-based docking models due to erratic and ineffective results, adopting flexible docking with GOLD instead.
Use of Computational Methods To Understand The Pattern Of Antimicrobial Resistance · Nottingham · PhD
Considered and rejected
Considered and rejected: Rejected rigid-body docking in favor of flexible docking with HDOCK/HADDOCK to allow atomic side-chain rotations and mutual conformational adaptation at large protein-protein interfaces.
El rol estructural de la subunidad β3 como regulador de hNav1.7 · BUAP · PhD
Considered and rejected
Considered and rejected: Rigid receptor docking was rejected in favor of flexible-residue docking within a 5 Å active-site radius to correctly capture aromatic ligand stabilization.
Computational approaches for mapping, understanding and modulating interactions in microbial communities · Boston University · PhD
Considered and rejected
Considered and rejected: Rejected rigid receptor docking in GOLD due to extreme steric clashes and negative fitness scores.
Experimental and computational study of latex clearing protein LcpK30 for rubber degradation · Nottingham · PhD
Considered and rejected
Considered and rejected: AutoDock Vina rigid-body docking rejected for variants containing active-site F87/bulky residues due to severe steric exclusion of productive poses; ADFR flexible-receptor docking used instead.
Enantioselective remote functionalisation of cyclic amines by engineered P450BM3 variants · Oxford · PhD
Considered and rejected
Considered and rejected: Constraining rotational bonds of the 90-residue EREG and TGFα substrate segments during AutoDock Vina docking to treat them as rigid bodies due to structural complexity.
Identification of Molecular Determinants in iRhoms1/2 and EGFR Ligands That Contribute to the Substrate Selectivity of Stimulated ADAM17 · Cornell · PhD
Docking scoring functions fail to correlate with experimental binding affinity and potency
Calculated docking scores and binding energies repeatedly failed to reproduce experimental rank orders, functional potencies, or biophysical assay hits. In addition, scoring metrics exhibited persistent background correlations at extreme docking errors and were susceptible to exploitation by generative models.
Tried and failed
graph theoretical energy scoring on docked poses applied to allosteric ligand affinity ranking. Reason: scoring accuracy degraded significantly due to docking pose inaccuracies and geometric deviations
Allostery and signalling pathways: methods and applications through graph theory · Imperial · PhD
Tried and failed
molecular docking virtual screening compared to biophysical assays applied to fragment library screening against RNA targets. Outcome: no signal. Reason: computational docking hits showed zero overlap with surface plasmon resonance experimental binding hits
Can In-Silico Computational Virtual Screening of Fragment Libraries Model In-Vitro Biophysical Screening Methods with Similar Results to Identify Novel RNA-Binding Chemical Matter · Harvard · Master's
Tried and failed
structure-based molecular docking score ranking applied to target enzyme ligand screening. Outcome: no signal. Reason: docking scores showed no correlation with known positive and negative control performance
Lost to a baseline
Molecular docking scores showed a weak correlation with functional potency (R2 < 0.1), failing to accurately predict cellular IC50 ranking.
Design, Synthesis, and Development of Novel Intracellular Allosteric Modulators of CXC Chemokine Receptor 1 and 2 · Nottingham · PhD
Lost to a baseline
On CASF 2016 docking RMSD bins (e.g. 0-1Å and 1-2Å), all MLBSFs and physics-based scoring plateaued at ~10Å docking error and retained non-zero baseline correlation even at extreme docking errors (25-30Å), matching non-structural ligand bias behavior.
Towards addressing structural data limitations in machine learning for small molecule drug discovery · Oxford · PhD
Considered and rejected
Considered and rejected: Scoring solely based on docking energy calculations was rejected because it fails to distinguish partial agonists from full or dual agonists.
Defining Novel Clusters of PPAR gamma Partial Agonists for Virtual Screening · Virginia Tech · Master's
Considered and rejected
Considered and rejected: Molecular docking scoring (AutoDock4) was rejected as a quantitative alternative/replacement for experimental G-quadruplex DNA binding affinity measurements due to inability to preserve true rank order and errors exceeding method uncertainty (2.5 kcal/mol)
Discovery of photoactive anti-cancer platinum(II) complexes: synthesis, screening and machine learning approaches · Imperial · PhD
Tried and failed
3D-CNN scoring of docked protein-ligand poses applied to allosteric inhibitor binding prediction. Outcome: did not generalise. Reason: Training solely on orthosteric active-site docked structures failed to generalize to allosteric binding modes
Leveraging diverse data modalities to study kinase inhibitor polypharmacology · Harvard · PhD
Tried and failed
reinforcement learning on docking score objective alone applied to de novo small molecule generation. Outcome: overfit. Reason: model exploited the scoring oracle, generating overly large, highly lipophilic molecules with poor drug-likeness
Language-based Sample-efficient and Synthesizable Molecular Generative Design · EPFL · PhD
Tried and failed
geometric deep learning for de novo ligand design applied to protein-ligand binding generation. Outcome: no signal. Reason: computationally designed compound failed to show measurable binding affinity in experimental assays
Computational Design of Structure-Guided Biomolecular Interactions · EPFL · PhD
Inaccurate receptor conformations and incorrect oligomeric assemblies prevent productive docking
Homology models and unliganded crystal structures frequently exhibited collapsed cavities or missing tunnels that physically precluded ligand entry. Similarly, attempting to dock against isolated monomer subunits failed to capture proper binding environments present only in intact multimeric complexes.
Tried and failed
3D CNN scoring of molecular docking poses applied to protein kinase ligand binding prediction. Outcome: did not generalise. Reason: experimental assays failed to confirm predicted inhibitor binding to homology-modeled targets
Leveraging diverse data modalities to study kinase inhibitor polypharmacology · Harvard · PhD
Tried and failed
molecular docking on homology-predicted protein structures applied to small molecule ligand-target binding identification. Outcome: no signal. Reason: predicted closed-conformation model failed to capture appropriate binding pockets for the ligand
Understanding responses to chemical and environmental perturbations in blood stage Babesia parasites · Harvard · PhD
Tried and failed
molecular docking of cofactor into active site applied to predicted oxidoreductase enzyme structures. Reason: severe steric clashes and absence of a substrate-binding tunnel prevented docking
Expanding the Tree of Life and altering models of eukaryogenesis · UT Austin · PhD
Tried and failed
molecular dynamics simulation of ligand binding applied to intact multi-protein complex. Outcome: did not generalise. Reason: isolated domain binding mode was not reproduced in the intact multi-subunit protein assembly
Atomistic simulations of cardiac troponin with phosphorylation andmutation effects · Imperial · PhD
Considered and rejected
Considered and rejected: Direct receptor-ligand docking of homotrimers was rejected because it produces an artificial dimeric state instead of a trimeric assembly, requiring symmetry docking instead.
Investigating determinants of tumor escape and developing computational tools for protein-protein interface analysis · Boston University · PhD
Considered and rejected
Considered and rejected: Rejected performing molecular docking between collagen and the monomeric NTHiENO model because contact residues mapped to internal dimer interfaces causing steric clashes, requiring a dimeric homology model.
Considered and rejected
Considered and rejected: Rejected building CXCR1 homology models in Modeller without the template-bound ligand, as unliganded models collapsed the binding cavity due to rotamer shifts of I64(1.57).
Design, Synthesis, and Development of Novel Intracellular Allosteric Modulators of CXC Chemokine Receptor 1 and 2 · Nottingham · PhD
Unguided and unrestrained search protocols produce distorted or nonconvergent poses
Macromolecular docking without structural restraints produced hundreds of divergent models lacking a consensus binding interface. Unconstrained searches also led to distorted bond geometries for flexible peptides and frequently yielded nonsensical poses when template guidance was omitted.
Considered and rejected
Considered and rejected: Rejected raw unguided HADDOCK docking without restraints because guided docking with ECLAIR predictions significantly improved interface recall (0.29 vs 0.21, p=5.88e-6) and structural RMSD (9.45 Å vs 11.79 Å, p=0.04).
Considered and rejected
Considered and rejected: Avoided de novo docking due to generation of non-sensical poses, relying instead on template-based docking.
Tried and failed
rigid-body protein-protein macromolecular docking applied to heteromeric enzyme complex modeling. Outcome: did not converge. Reason: produced over 400 highly variable models without a consensus interface
Structural investigations of adenosylcobalamin-dependent enzyme maturation · MIT · PhD
Tried and failed
direct flexible backbone docking with non-canonical amino acids applied to peptide-protein complex modeling. Outcome: unstable. Reason: produced distorted bond angles and lengths without initial canonical amino acid backbone pre-refinement
FROM SIMULATIONS TO LANGUAGE MODELS: COMPUTATIONAL INNOVATIONS FOR CHEMICAL BIOLOGY AND DRUG DISCOVERY · Penn · PhD
Considered and rejected
Considered and rejected: Rejected relying solely on minimum free-energy docking poses because long-range dipole-charge interactions predetermine orientation prior to binding.
FIRST PRINCIPLES STUDIES ON STRUCTURE-CARCINOGENICITY RELATION · Saskatchewan · PhD
Considered and rejected
Considered and rejected: Rejected predicting ligand binding affinities using an unguided whole-head search without domain decomposition because it lowers specificity and search accuracy.
Computational Modeling of Cell-Substrate Interactions · Johns Hopkins · PhD
Docked ligand poses drift away or bind to off-target locations
Ligands placed into targeted binding pockets frequently drifted away when subjected to molecular dynamics refinement or when geometric constraints were removed. In several instances, scoring functions preferred non-catalytic or off-target regions outside the intended active site.
Tried and failed
molecular docking with score thresholding applied to RNA-fragment binding site prediction. Reason: High-scoring conformers docked to off-target residues outside the targeted binding site
Can In-Silico Computational Virtual Screening of Fragment Libraries Model In-Vitro Biophysical Screening Methods with Similar Results to Identify Novel RNA-Binding Chemical Matter · Harvard · Master's
Tried and failed
constrained docking followed by unconstrained validation docking applied to enzyme active site mutant design. Outcome: did not generalise. Reason: predicted intermediate poses were unstable and drifted away from target catalytic sites when geometric constraints were removed
Computational and structural investigations of class I diterpene synthases · Iowa State · PhD
Left open by the authors
Problems the authors named and did not get to.
Left open
Simulate membrane deformation and endocytosis initiation driven by localized multivalent receptor-ligand accumulation in fluid membranes. Blocker: Lacks specific model parameters, coupling mechanism to deformable membrane models, and benchmark targets
Controlling and designing the binding properties of multivalent systems: a coarse-grained modelling and simulations approach · Imperial · PhD
Left open
Integrate propensity optimised paths with ensemble docking to evaluate multiple ligand poses in fragment-based virtual screening for allosteric drug design. Blocker: None
Allostery and signalling pathways: methods and applications through graph theory · Imperial · PhD
Left open
Integrate ligand pose scoring and drug-like property predictors into the LDDM generative pipeline to constrain hit identification. Blocker: None
Computational Design of Structure-Guided Biomolecular Interactions · EPFL · PhD
Left open
Dock diverse pathogen-derived lipid structures into primate CD1a structural models using molecular docking tools to test binding affinity. Blocker: None
Adaptive Evolution in Primate Immune Receptors · University of Oregon · PhD
Left open
Experimentally validate the computational docking and molecular dynamics model of the proteusin precursor peptide interacting with peptide brominase SrpI. Blocker: Requires wet-lab experimental validation (biochemical binding/activity assays)
Biosynthesis Guided the Discovery of Proteusin RiPP Natural Products · Georgia Tech · PhD
Left open
Finetune GeoMol on molecular docking pose generation or 4D QSAR descriptor prediction datasets. Blocker: None
Towards Automated Reaction Kinetics with Message Passing Neural Networks · MIT · PhD
Left open
Benchmark RoseTTAFold All-Atom on predicting ligand-induced protein conformational changes using public apo/holo paired datasets. Blocker: None
Explainable Machine Learning and Applications in Protein-Ligand Complex Structure Prediction · University of Washington · PhD
Who has worked on this
Authors of the public theses cited on this page, most-cited first.
- Pascal Sturmfels · University of Washington · PhD · 2024
Explainable Machine Learning and Applications in Protein-Ligand Complex Structure Prediction - Freyr Sverrisson · EPFL · PhD · 2024
Predicting protein interactions using geometric deep learning on protein surfaces - Irina Morozova · EPFL · PhD · 2026
Computational Design of Structure-Guided Biomolecular Interactions - G Durant · Oxford · PhD · 2025
Towards addressing structural data limitations in machine learning for small molecule drug discovery - Nan Wu · Imperial · PhD · 2024
Allostery and signalling pathways: methods and applications through graph theory - Rhys Francis · Nottingham · PhD · 2024
Design, Synthesis, and Development of Novel Intracellular Allosteric Modulators of CXC Chemokine Receptor 1 and 2 - Ameya Dhatri Madduri · Harvard · Master's · 2023
Can In-Silico Computational Virtual Screening of Fragment Libraries Model In-Vitro Biophysical Screening Methods with Similar Results to Identify Novel RNA-Binding Chemical Matter for Treatment of Myo - Changchang Liu · Harvard · PhD · 2021
Leveraging diverse data modalities to study kinase inhibitor polypharmacology - Ali Abdolahzadeh Ziabari · Saskatchewan · PhD · 2026
INTEGRATING ELECTRONIC STRUCTURE AND DYNAMIC STABILITY IN A PREDICTIVE MODEL FOR THE TOXICITY OF DIOXIN-LIKE COMPOUNDS - Molly Bartlett · Imperial · PhD · 2025
Discovery of photoactive anti-cancer platinum(II) complexes: synthesis, screening and machine learning approaches - Woody Ahern · University of Washington · PhD · 2025
Diffusion Models for Protein Structure Design: From Backbone Generation to Atomic-Resolution Enzyme Design - M.C. Jordan Edilberto Ruiz Castelán · BUAP · PhD · 2025
El rol estructural de la subunidad β3 como regulador de hNav1.7
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.