The Use of Deep Learning in RNA Therapeutic Development.

Area of Science:

• Biotechnology
• Genomics
• Computational Biology

Background:

• Ribonucleic acid (RNA)-based therapeutics offer a promising, versatile, and less toxic alternative to traditional gene therapies.
• The vast design space of RNA therapeutics, including codon identities and secondary structures, poses significant optimization challenges.
• Experimental exploration of this extensive design space is often impractical, necessitating advanced computational approaches.

Purpose of the Study:

• To review the application of deep learning (DL) models in optimizing RNA therapeutic development.
• To highlight DL's contributions to RNA structure prediction, CRISPR activity, and RNA delivery.
• To analyze the impact of DL model architectures and discuss future directions.

Main Methods:

• Review of existing literature on deep learning applications in RNA therapeutics.
• Analysis of deep learning model architectures and their performance in key areas.
• Discussion of computational and data limitations.

Main Results:

• Deep learning models have made significant contributions to RNA structure prediction, enhancing CRISPR activity, and improving RNA delivery strategies.
• Different deep learning model architectures show varying effects on performance in RNA therapeutic development.
• Computational and data limitations present key challenges for current deep learning applications.

Conclusions:

• Deep learning is a powerful tool for navigating the complex design space of RNA therapeutics.
• Future research should focus on emerging DL architectures and integration with high-throughput screening.
• Continued advancements in deep learning promise to further revolutionize RNA therapeutic development.