Skip to main content
recep.adiyaman
Vision About Experience Education Skills Publications Tools Newsletter Updates Podcasts Blog Contact CV Subscribe
All Issues
Older Newer
Daily Signal May 15, 2026 · 8 min read

Issue #109: AlphaFold and the Transformation of Structural Biology: Evolution, Applications, Limitations, and Future Directions

Protein Design Digest #109: AlphaFold and the Transformation of Structural Biology: Evolution, Appli…

Share X LinkedIn
Protein Design Daily

Building something in Protein Design?

I love collaborating on new challenges. Let's build together.

LinkedIn Email

Subscribe to Protein Design Digest

Daily curated signals from arXiv, PubMed, and BioRxiv.

Signal of the Day

AlphaFold and the Transformation of Structural Biology: Evolution, Applications, Limitations, and Future Directions

The protein folding problem is the challenge of predicting a protein’s three-dimensional structure from its amino acid sequence. This problem has been a central challenge in molecular biology for over fifty years. The advent of AlphaFold, a deep learning system developed by DeepMind, represented a paradigm shift in structural biology by demonstrating near-experimental accuracy in protein structure prediction. This review traces the evolution of the AlphaFold family of models, from its breakthrough performance in CASP14 through the expanded capabilities of AlphaFold 3 and the recent emergence of the proprietary Isomorphic Labs Drug Design Engine (IsoDDE). We examine the architectural innovations that underpin AlphaFold’s success, its broad applications in drug discovery, virology, and protein engineering, and its well-documented limitations in modeling intrinsically disordered regions, conformational ensembles, and allosteric mechanisms. We also discuss the growing tension between open science and commercial interests in AI-driven structural biology. The review draws primarily on peer-reviewed literature and curated expert sources to provide an accessible yet rigorous overview of the current state and future trajectory of AI-based protein structure prediction.

Why this matters: Critical for improving fold accuracy and reducing structural uncertainty in de novo design.

Also Worth Reading

PathDiffusion: modeling protein folding pathway using evolution-guided diffusion

Despite remarkable advances in protein structure prediction, a fundamental question remains unresolved: how do proteins fold from unfolded conformations into their native states? Here, we introduce PathDiffusion, a novel generative framework that simulates protein folding pathways using evolution-guided diffusion models. PathDiffusion first extracts structure-aware evolutionary information from 52 million predicted structures the AlphaFold database. Then an evolution-guided diffusion model with a dual-score fusion strategy is trained to generate high-fidelity folding pathways. Unlike existing deep learning methods, which primarily sample equilibrium ensembles, PathDiffusion explicitly models the temporal evolution of folding. On a benchmark of 52 proteins with experimentally validated folding pathways, PathDiffusion accurately reconstructs the order of folding events. We further demonstrate its versatility across four challenging applications: (1) recapitulating Antons molecular dynamics trajectory for 12 fast-folding proteins, (2) reproducing functionally important local folding-unfolding transitions in 20 proteins, (3) characterizing conformational ensembles of 50 intrinsically disordered proteins, and (4) resolving distinct folding mechanisms among 3 TIM-barrel proteins. We anticipate that PathDiffusion will be a valuable tool for probing protein folding mechanisms and dynamics at scale.

Integrated computational modeling of chalcone-based inhibitors targeting carbonic anhydrase I and II: 3D-QSAR, molecular docking, and dynamics simulations.

Carbonic anhydrase isoforms I and II (hCA-I and hCA-II) are metalloenzymes involved in essential physiological processes and represent relevant therapeutic targets for disorders such as glaucoma and osteoporosis. Chalcones have emerged as promising scaffolds for carbonic anhydrase inhibition; however, their structure-activity relationships, particularly for non-sulfonamide derivatives, remain insufficiently explored from a computational point of view. In this study, a dataset of 118 chalcone derivatives has been analyzed by using a three-dimensional quantitative structure-activity relationship (3D-QSAR) modeling, which comprises Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Index Analysis (CoMSIA). The developed models exhibited strong internal consistency and predictive capability for both isoforms. For hCA-I, steric, electrostatic, hydrophobic, and hydrogen bond acceptor fields has been identified as key contributors to inhibitory activity, whereas for hCA-II, hydrogen bond donor features played a more prominent role. Molecular docking and molecular dynamics simulations have been employed as complementary approaches to analyze ligand-protein interactions and binding stability. In addition, quantum chemical descriptors, derived from density functional theory, that have been integrated with the 3D-QSAR analysis, reveal a consistent correspondence between contour map features and the distribution of frontier molecular orbitals and molecular electrostatic potential. Furthermore, ADME-based pharmacokinetic properties of the proposed compounds have been evaluated to assess their potential drug-likeness. Based on the integrated computational analysis, six new chalcone derivatives, with predicted inhibitory activity in the nanomolar range, are proposed. Overall, this study provides a consistent physicochemical framework for understanding the inhibitory activity of chalcone derivatives and highlights key molecular features that may guide the modulation of activity across hCA-I and hCA-II isoforms.

Toward Accurate RNA Folding Thermodynamics: Evaluation of Enhanced Sampling Methods for Force Field Benchmarking.

Biologically functional RNAs operate near marginal stability, and their rugged free-energy landscapes and profound structural dynamics - typically not captured by structural biology experiments - play decisive roles. Atomistic molecular dynamics (MD) simulations provide a unique means to characterize these features. However, the applicability of atomistic MD is currently limited by accessible simulation time scales and, most importantly, by force-field (FF) accuracy. Folding free energies (ΔG°fold) of small RNA motifs represent well-defined targets for quantitative benchmarking of RNA FFs. In practice, however, obtaining thermodynamic estimates that are sufficiently robust for direct comparison with experimental data remains highly challenging, even for small RNA systems, and many published studies rely on sampling that is not fully converged. Here, we systematically assess the performance of widely used advanced enhanced sampling techniques using the 8-mer r(gcGAGAgc) tetraloop as a representative benchmark system. We test temperature replica exchange (T-REMD), two solute-tempering variants of replica exchange (REST2 and REHT), as well as well-tempered metadynamics and on-the-fly probability enhanced sampling combined with solute tempering (ST-MetaD and ST-OPES). Among the tested approaches, T-REMD proves to be the most robust, yielding reproducible folding equilibria and consistent estimates of ΔG°fold after approximately 20 μs of simulation time, independent of the initial folded or unfolded conformational ensemble. Our results provide practical guidelines for selecting sampling protocols suitable for quantitative RNA benchmarks and lay the foundation for systematic validation and future refinement of RNA FFs.

Research & AI Updates

From the Industry

Quick Reads

In silico identification of NPACT-derived PAK1 inhibitors using machine learning, molecular docking, and dynamic simulation approaches.

P21-activated kinase 1 (PAK1) is a key serine/threonine kinase involved in cytoskeletal remodeling, cell proliferation, and survival, and its aberrant activation has been strongly associated with tumorigenesis in multiple cancer types. Read more →

Identification of Novel Extracellular-Signal-Regulated Kinase 2 Inhibitors Through Machine Learning-Driven De Novo Design, Molecular Docking, and Free-Energy Perturbation.

Background : The extracellular-signal-regulated kinase (ERK) cascade regulates cell proliferation, differentiation, and survival, and ERK2 mediates substrate phosphorylation, influencing gene expression and cellular functions. Read more →

Natural compounds extracted from medicinal herbs in the treatment of Parkinson’s disease; Molecular Docking, Molecular Dynamics simulation, and Quantum Mechanical calculations.

The dopamine D3 receptor (D3R), which belongs to class A of G-protein-coupled receptors (GPCRs), is a promising target and is significantly involved in the pathology of Parkinson’s disease progression. Read more →

Investigating the Impact of Bisphenol A on Gastric Cancer Through Network Toxicology and Molecular Docking.

The extensive application of bisphenol A (BPA), an endocrine-disrupting substance in polymer-based products and consumer commodities, has sparked concerns over its carcinogenic potential. Read more →

Shoebill: an interpretable AlphaFold2-informed predictor of protein crystallization propensity using XGBoost.

X-ray crystallography is a central technique for high-resolution protein structure determination; however, the production of diffraction-quality crystals remains a major bottleneck due to high experimental cost and low success rates. Read more →

Exploring the mechanism of plastic additives on reproductive disorders based on network toxicology and molecular docking.

Phthalate esters (PAEs) and organophosphate esters (OPEs) are pervasive 20 environmental pollutants with endocrine-disrupting properties. Read more →

Nanoemulsion advancements for effective Sargassum serratifolium delivery in cosmetics: extraction, characterisation, and molecular docking.

In the cosmetic industry, active ingredients often consist of organic compounds that are hydrophobic, making oil-based formulations impractical. Read more →

LLM-Enhanced Knowledge Distillation for Sequence-Based Protein-Ligand Interaction Prediction.

Accurate prediction of protein-ligand interactions is essential for drug discovery, supporting critical stages from lead optimization to therapeutic development. Read more →

Pipeline Tip

Use local MSA generation (colabfold_search) to bypass speed bottlenecks.

Resources & Tools

Deep learning is not a magic wand, but a powerful lens for structural biology. — Recep Adiyaman

BS HF DK