Generative AI for Accurate DNA Binding Protein Prediction Using Computer Vision

Read full text Summarise with VIRAT
Authors: Meenakshi K, Sharmila C, Monanjana V, Senthil G A
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The accurate prediction of DNA-protein binding is important for understanding genetic regulation, drug discovery, and therapeutic design. Traditional computational models often have difficulty capturing the complex relationships between DNA sequences and binding proteins. To tackle this issue, the proposed study introduces Deep Gen Bind, a framework that uses generative artificial intelligence and computer vision techniques to predict DNA-protein interactions accurately. The approach combines algorithms like Generalized Support Vector Machine (GSVM), Generative Adversarial Networks (GAN), and Deep Convolution GANs (DCGAN) to model both sequence patterns and structural features. GSVM offers a solid baseline by classifying binding affinities with an accuracy of 89%. GAN-based models improve learning by creating realistic synthetic representations of DNA- protein complexes, reaching 93% accuracy. The DCGAN further boosts performance by extracting hierarchical features and capturing complex spatial dependencies, leading to a prediction accuracy of 97%. By using computer vision techniques on biological data, this research shows that deep generative models can overcome data limitations and improve generalization. The results highlight not just the potential of AI-driven generative methods in bio-informatics but also their advantages over traditional approaches. Ultimately, Deep Gen Bind aids in providing more reliable predictions of DNA- binding proteins, supporting advancements in molecular biology, precision medicine, and biotechnology applications.

Fields of Study & Subjects

Machine Learning in Bioinformatics Genomics and Chromatin Dynamics Protein Structure and Dynamics

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
22 March 2026
Language
EN
Document type
Book