Drug–Target Affinity Prediction Based on Cross-Modal Fusion of Text and Graph DOI Creative Commons
Jucheng Yang, Fenghui Ren

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(6), P. 2901 - 2901

Published: March 7, 2025

Drug–target affinity (DTA) prediction is a critical step in virtual screening and significantly accelerates drug development. However, existing deep learning-based methods relying on single-modal representations (e.g., text or graphs) struggle to fully capture the complex interactions between drugs targets. This study proposes CM-DTA, cross-modal feature fusion model that integrates textual molecular graphs with target protein amino acid sequences structural graphs, enhancing diversity expressiveness. The employs multi-perceptive neighborhood self-attention aggregation strategy first- second-order information, overcoming limitations graph isomorphism networks (GIN) for representation. experimental results Davis KIBA datasets show CM-DTA improves performance of drug–target prediction, achieving higher accuracy better metrics compared state-of-the-art (SOTA) models.

Language: Английский

Advancements in the integration of isothermal nucleic acid amplification methods for point-of-care testing in resource-limited settings. DOI Creative Commons
Noemi Bellassai, Roberta D’Agata, Giuseppe Spoto

et al.

Sensors and Actuators Reports, Journal Year: 2025, Volume and Issue: unknown, P. 100285 - 100285

Published: Jan. 1, 2025

Language: Английский

Citations

0
Promising Solutions to Address the Non-Specific Adsorption in Biosensors Based on Coupled Electrochemical-Surface Plasmon Resonance Detection DOI Creative Commons
Alina Vasilescu, Szilveszter Gáspár, Mihaela Gheorghiu

et al.

Chemosensors, Journal Year: 2025, Volume and Issue: 13(3), P. 92 - 92

Published: March 5, 2025

Nonspecific adsorption (NSA) impacts the performance of biosensors in complex samples. Coupled electrochemical–surface plasmon resonance (EC-SPR) offer interesting opportunities to evaluate NSA. This review details main solutions minimize fouling electrochemical (EC), surface (SPR) and EC-SPR biosensors. The discussion was centered on blood, serum milk as examples matrices. Emphasis placed antifouling coatings, NSA evaluation protocols universal functionalization strategies obtain In last 5 years, various coatings were developed for EC biosensors, including new peptides, cross-linked protein films hybrid materials. Due comparatively much more scarce literature, SPR extended early 2010s. analysis revealed a wide range materials with tunable conductivity, thickness functional groups that can be tested future EC-SPR. high-throughput screening materials, molecular simulations machine learning-assisted evaluations will even further widen available minimization NSA’s impact analytical signal is moreover facilitated by unique sensing mechanisms associated bioreceptor or particularities detection method. It hoped this encourage research field

Language: Английский

Citations

0
Drug–Target Affinity Prediction Based on Cross-Modal Fusion of Text and Graph DOI Creative Commons
Jucheng Yang, Fenghui Ren

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(6), P. 2901 - 2901

Published: March 7, 2025

Drug–target affinity (DTA) prediction is a critical step in virtual screening and significantly accelerates drug development. However, existing deep learning-based methods relying on single-modal representations (e.g., text or graphs) struggle to fully capture the complex interactions between drugs targets. This study proposes CM-DTA, cross-modal feature fusion model that integrates textual molecular graphs with target protein amino acid sequences structural graphs, enhancing diversity expressiveness. The employs multi-perceptive neighborhood self-attention aggregation strategy first- second-order information, overcoming limitations graph isomorphism networks (GIN) for representation. experimental results Davis KIBA datasets show CM-DTA improves performance of drug–target prediction, achieving higher accuracy better metrics compared state-of-the-art (SOTA) models.

Language: Английский

Citations

0