We have published a new part in the Genome Toolkit series: Part 4.6 – Building Structured Scientific Results.

In this part, we upgrade Genome Toolkit from raw string inputs and primitive return values to validated biological inputs and structured scientific results. Our k-mer algorithms now accept validated Sequence objects and return typed results containing metadata, inputs, parameters, and outputs.
We also build a more realistic experiment workflow by loading and validating one DNA sequence once, then reusing that same DNA object across multiple independent k-mer experiments. Each result keeps compact provenance without duplicating the full biological sequence, and we demonstrate both direct access to individual values and complete JSON serialization.

Article: https://rebelscience.club/2026/09/genome-toolkit-part-4-6-building-structured-scientific-results/

Video: https://youtu.be/8X9dAnpZUSs

We have published the final part of the Genome Toolkit Part 4 refactor: Part 4.7 – Testing, Reliable Errors, and Final Polish.

In this part, we add clear Genome Toolkit exception types and build our first automated test suite with pytest. We test our sequence models, plain-text and FASTA loaders, and k-mer algorithms, including both normal behavior and important error cases.

The tests also uncover several real edge cases in our existing k-mer algorithms. We fix lowercase k-mer handling, reject empty k-mers and invalid k-mer lengths clearly, and then run the complete test suite and final project verification.

With Part 4 complete, we can now return to bioinformatics and use this stronger Genome Toolkit foundation to add new algorithms, work with real biological data, and build more scientific experiments.

Article: https://rebelscience.club/2026/09/genome-toolkit-part-4-7-testing-reliable-errors-examples-and-final-polish/

Video: https://youtu.be/5cvMr2OFFcc