Biology subjects

Mey, A.

Publications and source records attributed to Mey, A..

8 recordsLinked to original sources

bioRxiv2025

Explainable Active Learning Framework for Ligand BindingAffinity Prediction

Active learning (AL) prioritises which compounds to measure next for protein-ligand affinity when assay or simulation budgets are limited. We present an explainable AL framework built on Gaussian process regression and assess how molecular representations, covariance kernels, and acquisition policies affect enrichment across four drug-relevant targets. Using recall of top active compounds, we find that dataset identity--the targets chemical landscape--sets the performance ceiling, while method choices modulate outcomes rather than overturn them. Fingerprints with simple Gaussian process kernels provide robust, low-variance enrichment, whereas learned embeddings with non-linear kernels can reach higher peaks but with greater variability. Uncertainty-guided acquisition consistently outperforms random selection, yet no single policy is universally optimal; the best choice follows structure-activity relationship (SAR) complexity. To enhance interpretability beyond black-box selection, we integrate SHapley Additive exPlanations (SHAP) to link high-impact fingerprint bits to chemically meaningful fragments across AL cycles, illustrating how the models attention progressively concentrates on SAR-relevant motifs.

biophysics↗

bioRxiv2025

IR spectroscopy: from experimental spectra to high-resolution structural analysis by integrating simulations and machine learning

Understanding biomolecular function at the atomic scale requires detailed insight into the structural changes underlying dynamic processes. Vibrational infrared (IR) spectroscopy--when paired with biomolecular simulations and quantum-chemical calculations--determines bond length variations on the order of 0.01 [A], providing insights into these structutral changes. Here, we address the forward problem in IR spectroscopy: predicting high-accuracy vibrational spectra from known molecular structures identified by biomolecular simulations. Solving this problem lays the groundwork for the inverse problem: inferring structural ensembles directly from experimental IR spectra. We evaluate two computational approaches, normal mode analysis and Fourier-transformed dipole autocorrelation, against experimental IR spectra of N-Methylacetamide, a prototypical model for peptide bond vibrations. Spectra are derived from simulation models at multiple levels of theory, including hybrid quantum mechanics/molecular mechanics, machine-learned and classical molecular mechanics approaches. Our results highlight the capabilities and limitations of current theoretical biophysical approaches to decode structural information from experimental vibrational spectroscopy data. These insights underscore the potential of future artificial intelligence (AI)-enhanced models to enable direct IR-based structure determination. For example, resolving the so far experimentally inaccessible structures of toxic oligomers involved in neurodegenerative diseases, enabling improved disease diagnostics and targeted therapies. TOC Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC="FIGDIR/small/665767v1_ufig1.gif" ALT="Figure 1"> View larger version (24K): org.highwire.dtl.DTLVardef@6603d3org.highwire.dtl.DTLVardef@1852c85org.highwire.dtl.DTLVardef@2ddaccorg.highwire.dtl.DTLVardef@77b554_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2023

Benchmarking active learning protocols for ligand binding affinity prediction

Active learning (AL) has become a powerful tool in computational drug discovery, enabling the identification of top binders from vast molecular libraries with reduced costs for relative binding free energy calculations and experiments. To design a robust AL protocol, it is important to understand the influence of AL parameters, as well as the features of the datasets on the outcomes. We use four affinity datasets for different targets (TYK2, USP7, D2R, Mpro) to systematically evaluate the performance of machine learning models (Gaussian Process model, Chemprop), sample selection protocols, as well as the batch size based on metrics describing the overall predictive power of the model (R2, Spearman rank, RMSE) as well as the accurate identification of top 2% / 5% binders (Recall, F1 score). Both models have a comparable Recall of top binders on large datasets, but the GP models surpass Chemprop when training data is sparse. A larger initial batch size, especially on diverse datasets, increased the Recall of both models as well as overall correlation metrics. However, for subsequent cycles, smaller batch sizes of 20 or 30 compounds proved to be desirable. Furthermore, the presence of Gaussian noise to the data, up to a certain threshold, still allowed the model to identify clusters with top-scoring compounds. However, excessive noise (<1{sigma}) did impact the models predictive and exploitative capabilities. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=111 SRC="FIGDIR/small/568570v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@cefb11org.highwire.dtl.DTLVardef@c522bcorg.highwire.dtl.DTLVardef@6b8f18org.highwire.dtl.DTLVardef@17f7706_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOTOC GraphicC_FLOATNO C_FIG

bioinformatics↗

bioRxiv2023

mRNA interactions with disordered regions control protein activity

The cytoplasm is compartmentalized into different translation environments. mRNAs use their 3'UTRs to localize to distinct cytoplasmic compartments, including TIS granules (TGs). Many transcription factors, including MYC, are translated in TGs. It was shown that translation of proteins in TGs enables the formation of protein complexes that cannot be established when these proteins are translated in the cytosol, but the mechanism is poorly understood. Here we show that MYC protein complexes that involve binding to the intrinsically disordered region (IDR) of MYC are only formed when MYC is translated in TGs. TG-dependent protein complexes require TG-enriched mRNAs for assembly. These mRNAs bind to a new and widespread RNA-binding domain in neutral or negatively charged IDRs in several transcription factors, including MYC. RNA-IDR interaction changes the conformational ensemble of the IDR, enabling the formation of MYC protein complexes that act in the nucleus and control functions that cannot be accomplished by cytosolically-translated MYC. We propose that certain mRNAs have IDR chaperone activity as they control IDR conformations. In addition to post-translational modifications, we found a novel mode of protein activity regulation. Since RNA-IDR interactions are prevalent, we suggest that mRNA-dependent control of protein functional states is widespread.

molecular biology↗

bioRxiv2021

Structural dynamics of the β-coronavirus Mpro protease ligand binding sites

{beta}-coronaviruses alone have been responsible for three major global outbreaks in the 21st century. The current crisis has led to an urgent requirement to develop therapeutics. Even though a number of vaccines are available, alternative strategies targeting essential viral components are required as a back-up against the emergence of lethal viral variants. One such target is the main protease (Mpro) that plays an indispensible role in viral replication. The availability of over 270 Mpro X-ray structures in complex with inhibitors provides unique insights into ligand-protein interactions. Herein, we provide a comprehensive comparison of all non-redundant ligand-binding sites available for SARS-CoV2, SARS-CoV and MERS-CoV Mpro. Extensive adaptive sampling has been used to explore conformational dynamics employing convolutional variational auto encoder-based deep learning, and investigates structural conservation of the ligand binding sites using Markov state models across {beta}-coronavirus homologs. Our results indicate that not all ligand-binding sites are dynamically conserved despite high sequence and structural conservation across {beta}-coronavirus homologs. This highlights the complexity in targeting all three Mpro enzymes with a single pan inhibitor.

biophysics↗

bioRxiv2018

Geometric constraints in protein folding

The intricate three-dimensional geometries of protein tertiary structures underlie protein function and emerge through a folding process from one-dimensional chains of amino acids. The exact spatial sequence and con1guration of amino acids, the biochemical environment and the temporal sequence of distinct interactions yield a complex folding process that cannot yet be easily tracked for all proteins. To gain qualitative insights into the fundamental mechanisms behind the folding dynamics and generic features of the folded structure, we propose a simple model of structure formation that takes into account only fundamental geometric constraints and otherwise assumes randomly paired connections. We find that despite its simplicity, the model results in a network ensemble consistent with key overall features of the ensemble of Protein Residue Networks we obtained from more than 1000 biological protein geometries as available through the Protein Data Base. Speci1cally, the distribution of the number of interaction neighbors a unit (amino acid) has, the scaling of the structures spatial extent with chain length, the eigenvalue spectrum and the scaling of the smallest relaxation time with chain length are all consistent between model and real proteins. These results indicate that geometric constraints alone may already account for a number of generic features of protein tertiary structures.\n\nAuthor summaryHow proteins fold constitutes one of the most persistent, broad, and exciting open research questions at the intersection of biology, chemistry, and physics. Which mechanisms induce a one-dimensional sequence of amino acids to form into a complex three-dimensional (3D) structure? Proteins in their active 3D structure impact most of the basic processes inside cells, including gene regulation, cell metabolism, and the creation of protein structures themselves. Yet, a general rule about which conditions lead to which speci1c 3D protein structures remains unknown to date.\n\nHere, we demonstrate how a simple model that takes only fundamental geometric constraints into account and otherwise assumes randomly paired connections, naturally generates an ensemble of folded structures that exhibits many of its coarse scale features consistent with those of protein residue networks resulting from tertiary structures of biological proteins. Speci1cally, we tested a set of more than 1000 biological proteins and model structures and extracted a range of ensemble properties, including the spatial extension with chain size, the distribution of the number of interacting neighbors in the folded structure, the spectrum of Laplacian eigenvalues, and the distribution of the dominant non-trivial eigenvalue. We found that all of those properties are consistent between the ensemble of biological protein residue networks and the networks emerging in a self-organized way from the simple model.\n\nThese results indicate that coarse ensemble properties of 3D protein structures are already induced by geometric constraints alone such that only finer scales of the folded structures of individual proteins are speci1cally controlled by the details of their amino acid sequences. Such simple models provide a new angle of analyzing protein structures at the coarse scale of ensembles and may help understand core mechanisms underlying the complex folding process.

biophysics↗

bioRxiv2018

Atomic Level Characterisation of Millisecond-Time Scale Protein Motions through a Combined Molecular Simulations and NMR Approach

Proteins need to interconvert between many conformations in order to function, many of which are formed transiently, and sparsely populated. Particularly when the lifetimes of these states approach the millisecond timescale, identifying the relevant structures and the mechanism by which they inter-convert remains a tremendous challenge. Here we introduce a novel combination of accelerated MD (aMD) simulations and Markov State modelling (MSM) to explore these excited conformational states. Applying this to the highly dynamic protein CypA, a protein involved in immune response and associated with HIV infection, we identify five principally populated conformational states and the atomistic mechanism by which they interconvert. A rational design strategy predicted that the mutant D66A should stabilise the minor conformations and substantially alter the dynamics whereas the similar mutant H70A should leave the landscape broadly unchanged. These predictions are confirmed using CPMG and R1{rho} solution state NMR measurements. By accurately and reliably exploring functionally relevant, but sparsely populated conformations with milli-second lifetimes in silico, our aMD/MSM method has tremendous promise for the design of dynamic protein free energy landscapes for both protein engineering and drug discovery.

biophysics↗

bioRxiv2017

Allosteric effects in a catalytically impaired variant of the enzyme Cyclophilin A are unrelated to millisecond time scale motions

There is much debate about the mechanisms by which molecular motions influence catalysis in enzymes. This work investigates the connection between stochastic protein dynamics and function for the enzyme cyclophilin A (CypA) in wild-type (WT) form, and three variants that features several mutations that are distal from the active site. Previous biophysical studies have suggested that conformational exchange between a major active and a minor inactive state on millisecond time scales plays a key role in catalysis for CypA. Here this hypothesis was addressed by a variety of molecular dynamic (MD) simulation techniques. The simulations reproduce X-ray crystallography derived evidence for a shift in populations of major and minor active site conformations between the wild-type and mutant forms. Strikingly, exchange between these active site conformations occurs at a rate that is 5 to 6 orders of magnitude faster than previously proposed. Further analyses indicate that the minor active site conformation is catalytically impaired, and that decreased catalytic activity of the mutants may be explained by changes in Phe113 motions on a ns-s time scale. Therefore previously described millisecond time scale motions may not be necessary to explain allosteric effects in CypA mutants.

biophysics↗