Rise of Deep Learning Clinical Applications and Challenges in Omics Data: A Systematic Review.

Area of Science:

• Bioinformatics
• Computational Biology
• Artificial Intelligence

Background:

• Deep learning (DL) shows promise for analyzing complex omics data.
• Existing literature reviews often lack a comprehensive view of DL in omics.
• Understanding DL's potential and challenges in omics is crucial for advancing biological research.

Purpose of the Study:

• To systematically review and evaluate scientific studies on deep learning models in the omics field.
• To demonstrate the potential of DL techniques in omics data analysis.
• To identify key challenges and provide guidelines for practitioners.

Main Methods:

• Systematic literature search conducted on IEEE Xplore, Web of Science, ScienceDirect, and PubMed from 2018-2022.
• Inclusion and exclusion criteria were applied to a total of 65 selected articles.
• Categorization of studies based on clinical applications, review papers, and comparative analysis/guidelines.

Main Results:

• 42 out of 65 articles focused on clinical applications of DL in omics data.
• 16 articles were review publications covering single- and multi-omics data.
• Only 7 articles focused on comparative analysis and guidelines.

Conclusions:

• Deep learning application in omics faces challenges in DL models, preprocessing, datasets, validation, and applications.
• Further research is needed to address these obstacles for effective DL implementation in omics.
• This study offers a valuable guideline for understanding DL's role in omics data analysis.