Demystifying dimensionality reduction techniques in the 'omics' era: A practical approach for biological science...

Area of Science:

• Biochemistry
• Molecular Biology
• Bioinformatics

Background:

• Dimensionality reduction is crucial for analyzing large 'omics' datasets.
• Common methods like PCA, t-SNE, and UMAP can be mathematically challenging for students.
• There is a need for accessible teaching methods for these complex techniques.

Purpose of the Study:

• To develop intuitive examples for teaching dimensionality reduction techniques.
• To demystify methods like PCA, t-SNE, and UMAP for students without strong mathematical backgrounds.
• To assess the impact of interactive teaching using COVID-19 data on student learning and interest.

Main Methods:

• Created intuitive examples using COVID-19 data for dimensionality reduction.
• Conducted a 4-hour practical session for postgraduate students using Python and Jupyter notebooks.
• Assessed learning impact through an anonymous survey.

Main Results:

• 67% of students found computer use enriched their learning.
• 66% valued Jupyter notebooks; 60% showed increased interest in Python.
• 40% reported increased interest and understanding of dimensionality reduction; 40% gained research skills.

Conclusions:

• Interactive teaching with intuitive examples and tools like Jupyter notebooks can effectively demystify dimensionality reduction.
• This approach can enhance student understanding, interest, and practical skills in bioinformatics.
• The developed examples serve as valuable resources for interactive teaching environments.