Biology subjects

Yesylevskyy, S. O.

Publications and source records attributed to Yesylevskyy, S. O..

4 recordsLinked to original sources

bioRxiv2026

Polymer Polydispersity and Lipid Composition Control Nanoplastic Disassembly in Membranes

Nanoplastic interactions with cell membranes can influence particle uptake, translocation, and biological effects, making their molecular description important for assessing the environmental and biological consequences of plastic pollution. However, molecular simulations generally represent nanoplastics using monodisperse polymer chains, whereas nanoplastics formed by environmental degradation are polydisperse. Here, we construct 4 nm polyethylene terephthalate (PET), polyethylene (PE), and polystyrene (PS) nanoparticles with lognormal molecular-weight distributions reported for degraded polymers, and with melt-like chain packing and entanglement. Using Martini coarse-grained molecular dynamics, we simulate their interactions for 10 microseconds with three lipid bilayers of increasing complexity. PE and PS insert rapidly and remain largely intact, whereas PET remains surface-associated and progressively releases individual chains. Fluid, polyunsaturated lipid environments undergo greater deformation and promote substantially more chain release, up to 56% of the chains for PET and 43% for PS. Release begins with the shortest chains and progressively extends to longer chains, while semicrystalline PE remains largely intact. In contrast, a size-matched monodisperse PS particle releases no chains and minimally perturbs the membrane. These results identify the molecular-weight distribution as a key determinant of nanoplastic-membrane interactions, and indicate that uniform model particles underestimate the release of polymer chains into biological membranes.

biophysics↗

bioRxiv2026

Critical Assessment of ML models for ADMET Prediction in TDC leaderboards

In this work we performed a critical assessment of the benchmarking procedures used in Therapeutics Data Commons (TDC) ADMET leaderboards, focusing on reproducibility, robustness against data leakage, and signs of test-set overfitting across all 22 TDC ADMET endpoints. For each endpoint, the top 3 leaderboard models were screened with a unified protocol: execution environment reproducibility check, data leakage assessment, verification of hyperparameter optimisation practices, and final re-evaluation of results and TDC ranking. Only 3 methods (CaliciBoost, MapLight, MapLight+GNN) passed all checks and showed overall reproducible performance, whereas most of top-ranked models exhibited unavailable code, non-reproducible execution environments, runtime incompatibilities, or various methodological flaws. In particular, we identified direct or indirect data leakages in MiniMol, GradientBoost and XGBoost models. We also used our in-house models based on the Mol2Vec architecture to investigate the consequences of deliberately overfitting on the TDC test set. It is shown that deliberate or accidental tuning on the public test set may lead to significant inflation of the model metrics and leaderboard position. Our results emphasize the urgent need for better public ADMET benchmarks with the hidden test sets, strict dataset versioning and model submission with standardized inference environments.

biophysics↗

bioRxiv2023

Boosting performance of generative diffusion model for molecular docking by training on artificial binding pockets

This study introduces the PocketCFDM generative diffusion model, aimed at improving the prediction of small molecule poses in the protein binding pockets. The model utilizes a novel data augmentation technique, involving the creation of numerous artificial binding pockets that mimic the statistical patterns of non-bond interactions found in actual protein-ligand complexes. An algorithmic method was developed to assess and replicate these interaction patterns in the artificial binding pockets built around small molecule conformers. It is shown that the integration of artificial binding pockets into the training process significantly enhanced the models performance. Notably, PocketCFDM surpassed DiffDock in terms of non-bond interaction quality, number of steric clashes, and inference speed. Future developments and optimizations of the model are discussed. AvailabilityThe inference code and final model weights of PocketCFDM are accessible publicly via the GitHub repository: https://github.com/vtarasv/pocket-cfdm.git.

bioinformatics↗

bioRxiv2020

TTAPE-Me dye is not selective to cardiolipin and responds to main anionic phospholipids unspecifically

Identification, visualization and quantitation of cardiolipin (CL) in biological membranes is of great interest due to important structural and physiological roles of this lipid. Selective fluorescent detection of CL using non-covalently bound fluorophore TTAPE-Me (1,1,2,2-tetrakis[4-(2-trimethylammonioethoxy)-phenylethene) has been recently proposed. However, this dye was only tested on wild-type mitochondria or liposomes containing neglegible amounts of other anionic lipids, such as PG and PS. No clear preference of TTAPE-Me for binding to CL compared to PG and PS was found in our experiments. The shapes of the emission spectra for these anionic phospholipids were also found to be indistinguishable. Our experiments and complementary molecular dynamics simulations suggest that fluorescence intensity of TTAPE-Me is regulated by dynamic equilibrium between emitting dye, bound to anionic lipids by means of unspecific electrostatic attraction, and non-emitting dye aggregates in aqueous solution. Therefore, TTAPE-Me is not suitable for detection, visualization and localization of CL in the presence of PS and PG present in physiological amounts in the membranes of eukaryotic and prokaryotic cells, respectively.

biochemistry↗