Betül Yüksel

Ms. Betül Yüksel


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 August 2025 in Biomolecules
Verified
Journal Article
Open access
PDF file
YSYusuf Buğra Severoğlu
Yusuf Buğra Severoğlu
BYBetül Yüksel
Betül Yüksel
CSCagatay Sucu
Cagatay Sucu
NSNeşe Aral Sözener
Neşe Aral Sözener
Vladimir N. UverskyVladimir N. Uversky
Vladimir N. Uversky
OCOrkid Coskuner-Weber
Orkid Coskuner-Weber
Solvents represent the quiet majority in biomolecular systems, yet modeling their influence with both speed and ri:gor remains a central challenge. This study maps the state of the art in implicit solvent theory and practice, spanning classical continuum electrostatics (PB/GB; DelPhi, APBS), modern nonpolar and cavity/dispersion treatments, and quantum–continuum models (PCM, COSMO/COSMO-RS, SMx/SMD). We highlight where these methods excel and where they falter, namely, around ion specificity, heterogeneous interfaces, entropic effects, and parameter sensitivity. We then spotlight two fast-moving frontiers that raise both accuracy and throughput: machine learning-augmented approaches that serve as PB-accurate surrogates, learn solvent-averaged potentials for MD, or supply residual corrections to GB/PB baselines, and quantum-centric workflows that couple continuum solvation methods, such as IEF-PCM, to sampling on real quantum hardware, pointing toward realistic solution-phase electronic structures at emerging scales. Applications across protein–ligand binding, nucleic acids, and intrinsically disordered proteins illustrate how implicit models enable rapid hypothesis testing, large design sweeps, and long-time sampling. Our perspective argues for hybridization as a best practice, meaning continuum cores refined by improved physics, such as multipolar water, ML correctors with uncertainty quantification and active learning, and quantum–continuum modules for chemically demanding steps.

31 July 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
YSYusuf Buğra Severoğlu
Yusuf Buğra Severoğlu
BYBetül Yüksel
Betül Yüksel
CSCagatay Sucu
Cagatay Sucu
NSNeşe Aral Sözener
Neşe Aral Sözener
OCOrkid Coskuner-Weber
Orkid Coskuner-Weber
Vladimir N. UverskyVladimir N. Uversky
Vladimir N. Uversky
Solvation plays a critical role in determining the structural, dynamic, and functional behavior of biomolecules. Implicit solvent models, which approximate the solvent as a continuum dielectric medium, offer computationally efficient alternatives to explicit solvent simulations while capturing key aspects of solvation energetics. This review presents a comprehensive overview of implicit solvent models, spanning classical electrostatic formulations such as the Poisson–Boltzmann and Generalized Born equations, to quantum-based continuum approaches including PCM, COSMO, and the SMx family. We explore the theoretical underpinnings of these models, their treatment of polar and nonpolar contributions to solvation free energy, and extensions such as hybrid methods and machine learning-augmented approaches. Special attention is given to modern developments like AGBNP2, ABSINTH, GBNSR6, and COSMO-RS, as well as recent advances in modeling intrinsically disordered proteins (IDPs), protein–ligand interactions, and nucleic acid dynamics. Applications in drug discovery, molecular biophysics, and electronic structure prediction are highlighted. We also discuss current limitations, including the treatment of solvent entropy, ion-specific effects, and parameter sensitivity, and propose future directions involving AI integration and multiscale modeling. By critically evaluating implicit solvent models across theoretical and applied dimensions, this work aims to guide in selecting appropriate frameworks for simulating solvation phenomena in biological systems.