Protein Sequence Analysis landscape: A Systematic Review of Task Types, Databases, Datasets, Word Embeddings Methods,...

Area of Science:

• Computational Biology
• Bioinformatics
• Artificial Intelligence in Biology

Background:

• Protein sequence analysis is crucial for understanding biological processes and genetic disorders.
• Wet-lab experiments for protein analysis are costly, time-consuming, and error-prone.
• There's a need to integrate Artificial Intelligence (AI) into large-scale proteomics for efficient analysis.

Purpose of the Study:

• To provide a comprehensive overview of AI-driven applications in protein sequence analysis.
• To bridge the gap between the fields of proteomics and AI for researchers.
• To facilitate the development of AI-driven protein sequence analysis tools.

Main Methods:

• Compilation of 63 distinct protein sequence analysis tasks.
• Detailed information on 68 protein databases and 627 benchmark datasets.
• Analysis of 25 word embedding methods and 13 language models used in AI applications.

Main Results:

• A comprehensive catalog of AI applications across 63 protein sequence analysis tasks.
• Facilitation of biological understanding for AI researchers.
• Identification of state-of-the-art performances for 63 tasks.

Conclusions:

• This work offers a holistic view of AI in protein sequence analysis, accelerating tool development.
• It serves as a valuable resource for both AI and proteomics researchers.
• The study highlights the potential of AI to revolutionize biological data analysis.