Representing DNA for machine learning algorithms: A primer on one‐hot, binary, and integer encodings
Biochemistry and Molecular Biology Education · 2025-03 · 53 (2) · 142-146
DOI: 10.1002/bmb.21870 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biochemistry and Molecular Biology Education, online erschienen am 05.12.2024. Verfasst von Yash Munnalal Gupta und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biology, Faculty of Science, Naresuan University, Phitsanulok, Thailand. Hauptthemen laut MEDLINE: Computational Biology, Machine Learning, DNA, Algorithms. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 17 Literaturangaben. Gefördert von: Naresuan University (NU), and National Science, Research and Innovation Fund (NSRF).
Abstract (Originalsprache)
This short paper presents an educational approach to teaching three popular methods for encoding DNA sequences: one‐hot encoding, binary encoding, and integer encoding. Aimed at bioinformatics and computational biology students, our learning intervention focuses on developing practical skills in implementing these essential techniques for efficient representation and analysis of genetic data. The primary goal of this study is to enhance students' understanding and practical application of DNA encoding methods, which are crucial for various computational analyses in bioinformatics. Our intervention consists of three key components: (1) a conceptual framework that contextualizes these encoding methods within broader bioinformatics applications, (2) an interactive Jupyter Notebook with Python code examples ( https://github.com/yashmgupta/Representing-DNA/tree/main ), and (3) a user‐friendly Streamlit application for visualizing encoded sequences ( https://dnaencoding.streamlit.app/ ) that also enables students to input their own DNA sequences and visualize the different encoding methods, further enhancing their understanding and practical experience. By combining conceptual overview with practical coding and visualization tools, our approach provides a comprehensive foundation for students to leverage these key DNA sequence encoding methods in their future work. This study contributes to bioinformatics education by offering effective, hands‐on learning resources that bridge the gap between theoretical knowledge and practical application in DNA sequence analysis, preparing students for advanced research and data analysis projects in the field.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Biology, Faculty of Science, Naresuan University, Phitsanulok, Thailand.
- Center of Excellence for Innovation and Technology for Detection and Advanced Materials (ITDAM), Naresuan University, Phitsanulok, Thailand.
- Business Management and Languages, Faculty of Management Science, Silpakorn University, Phetchaburi, Thailand.
ORCID-Profile: Yash Munnalal Gupta
Förderung
- Naresuan University (NU), and National Science, Research and Innovation Fund (NSRF)
Zitieren
APA
Gupta, Y. M., Kirana, S. N., & Homchan, S. (2025). Representing DNA for machine learning algorithms: A primer on one‐hot, binary, and integer encodings. Biochemistry and Molecular Biology Education, 53(2), 142-146. https://doi.org/10.1002/bmb.21870
Vancouver
Gupta YM, Kirana SN, Homchan S. Representing DNA for machine learning algorithms: A primer on one‐hot, binary, and integer encodings. Biochemistry and Molecular Biology Education. 2025;53(2):142-146. doi:10.1002/bmb.21870
BibTeX
@article{Gupta2025,
author = {Gupta, Yash Munnalal and Kirana, Satwika Nindya and Homchan, Somjit},
title = {Representing DNA for machine learning algorithms: A primer on one‐hot, binary, and integer encodings},
journal = {Biochemistry and Molecular Biology Education},
year = {2025},
volume = {53},
number = {2},
pages = {142--146},
doi = {10.1002/bmb.21870}
}
EUVIMED