Python for Biological Data Exploration and Visualization

Session 3 – 03:00:00 – introducing Seaborn
This session shifts the focus from data analysis to data visualization, emphasizing visual exploration as a tool for understanding data. Participants will begin with an overview of the Seaborn library and then explore core plot types used to examine distributions and relationships in data. The session covers histograms, kernel density estimates, and scatter plots, as well as advanced alternatives like hexbin and contour plots, which are helpful when working with large datasets.

Participants will also learn how to leverage Seaborn’s ability to map dataframe columns to visual elements such as color, size, and shape, and how to efficiently generate small multiple plots for deeper insights. As with pandas, the goal is to produce powerful, publication-quality visualizations with minimal and intuitive code.

Key Topics:
Seaborn syntax, distribution plots, relational plots, hexbin plots, contour plots, aesthetic mappings, small multiples, figure customization.

Session 4 – 03:00:00 – Categorical Axes with Seaborn
This session explores the wide range of visualization techniques available for categorical data in Seaborn. Participants will work with common plot types such as strip plots, box plots, and bar plots, as well as more advanced options including swarm plots, violin plots, and boxen plots. The variety of available chart types provides an opportunity to consider the trade-offs between simplicity and detail in data visualization.
The session also introduces techniques for customizing chart appearance, with an emphasis on effective and intentional use of color. Participants will review best practices for color selection and learn how to avoid common pitfalls that can reduce the clarity or impact of a visual.

Key Topics:
Categorical plots, strip plots, box plots, violin plots, swarm plots, boxen plots, color palettes, style customization, visual encoding, color best practices.

Session 5 – 03:00:00 – Grouping and Categories with pandas
With a foundation in data visualization established, this session returns to pandas to explore more advanced data manipulation techniques. Participants begin with an in-depth look at categorical data types and their role in organizing and interpreting datasets. The core focus of the session is the powerful groupby functionality in pandas, which enables aggregation and analysis across different groupings.

Participants will learn multiple strategies for grouping data, including grouping by existing columns, applying custom grouping functions, and creating bins from continuous values. Binning is introduced as a particularly useful method for transforming numerical data into categorical form, enabling richer comparisons and more effective visualizations.

Key Topics:
Categorical data types, grouping with groupby, aggregation, custom groupings, binning numerical data, categorical transformations, integration with visualizations.

Session 6 – 03:00:00 – Long vs. Wide Form Data and Heatmaps
This session introduces the concepts of long-form (tidy) and wide-form (summary) data, building on earlier pandas techniques for reshaping DataFrames. Participants will learn how and when to use each format effectively, recognizing that different forms serve different purposes in data analysis and visualization.

The session also introduces the final major chart type in Seaborn: the heatmap, along with its extension, the clustermap. These visualizations directly reflect the structure of summary tables, offering a compact and powerful way to detect patterns, highlight relationships, and explore high-dimensional data. Participants will apply heatmaps to practical examples that illustrate their value in solving complex visualization challenges.

Key Topics:
Long-form vs. wide-form data, data reshaping, pivoting and melting in pandas, heatmaps, clustermaps, matrix-style visualizations, interpreting color-coded data.

Session 7- 03:00:00 – Dictionaries
This session introduces Python’s dictionary data structure as a powerful alternative to lists for managing paired data. Using a bioinformatics example focused on k-mer counting, participants will explore why dictionaries are better suited for specific tasks and learn the syntax for creating, accessing, and manipulating dictionaries.

Comparisons between list-based and dictionary-based solutions highlight the strengths and appropriate use cases for each approach. Additional examples demonstrate how key-value data structures are common in both bioinformatics and general programming. Practical exercises include working with dictionaries in real-world contexts such as DNA-to-protein translation.

Key Topics:
Paired data structures, hashing, key uniqueness, argument unpacking, tuples, dictionary methods.

Session 8- 03:00:00 – File Management and Housekeeping Scripts
This session explores Python’s standard library modules for file manipulation, emphasizing automation of essential housekeeping tasks commonly encountered in bioinformatics projects. Participants will learn techniques for renaming, moving, deleting files, and creating directories programmatically.

The practical component focuses on a bioinformatics data preprocessing challenge: organizing DNA sequences by length. This exercise highlights important considerations such as managing program state across multiple runs, processing multiple input files, and generating multiple outputs efficiently.

Key Topics:
File input/output, directory management, file renaming and deletion, scripting automation, handling multiple files, program state persistence.