کنگره بین المللی زیست فناوری گیاهان دارویی و قارچ های کوهی

صفحه اصلی / پنجمین کنگره بین المللی و ششمین همایش ملی زیست فناوری گیاهان دارویی و قارچ های کوهی

Deep Learning and Multi-Omics Integration for Plant Secondary Metabolic Pathway Prediction

نویسندگان :

Zahra Sadeghi¹

1- دانشگاه سراوان

کلمات کلیدی :

Bioinformatics،Metabolomics،Natural Product Discovery،Pathway Engineering

چکیده :

Phytochemical pathway discovery is challenged by the complexity of plant metabolic networks, hindering the development of new natural products. To address this, we comprehensively reviewed databases like PubMed, Google Scholar, and Scopus to examine computational frameworks that integrate deep learning with multi-omics data. The results identified three primary strategies for data integration. Early integration combines all data into a single matrix, which is effective for straightforward prediction but can be challenged by high dimensionality. Intermediate integration seeks a shared, lower-dimensional representation of the data, making it a powerful tool for biological discovery by revealing latent relationships. Late integration analyzes each data type separately before merging the results, a method particularly useful in clinical applications where interpretability is crucial. While these approaches offer a more holistic understanding, they face challenges such as high data dimensionality and heterogeneity, with the most effective strategy ultimately depending on the specific research objective.