Colab Notebooks

These notebooks capture the larger, end-to-end strictosidine case study workflows used with teemi. They are the best starting point if you want to run the published DBTL examples in Google Colab with as little setup as possible.

Each notebook links to both Google Colab and the GitHub source file.

Strictosidine Case: First DBTL Cycle

Design

ID	Focus	Links
00	Automatically fetch homologs from NCBI in a standardizable and repeatable way.	Open in Colab | GitHub
01	Select promoters from RNA-seq data and fetch them from online databases with quality checks.	Open in Colab | GitHub
02	Generate a combinatorial library with DesignAssembly and prepare it for downstream build planning.	Open in Colab | GitHub

Build

ID	Focus	Links
03	Assemble a CRISPR plasmid with USER cloning.	Open in Colab | GitHub
04	Construct the background strain by knocking out G8H and CPR.	Open in Colab | GitHub
05	Generate the first combinatorial library covering 1280 possible combinations.	Open in Colab | GitHub

Test

ID	Focus	Links
06	Process LC-MS data and genotype the generated strains.	Open in Colab | GitHub

Learn

ID	Focus	Links
07	Use AutoML to predict the best combinations for a targeted second round of library construction.	Open in Colab | GitHub

Strictosidine Case: Second DBTL Cycle

Design

ID	Focus	Links
08	Translate machine-learning results into a targeted library of strains.	Open in Colab | GitHub

Build

ID	Focus	Links
09	Construct the targeted second-round library of strains.	Open in Colab | GitHub

Test

ID	Focus	Links
10	Process second-cycle LC-MS characterization data.	Open in Colab | GitHub

Learn

ID	Focus	Links
11	Run the second ML cycle and inspect improved performance and saturation of top strains.	Open in Colab | GitHub

Publication

For the full scientific context, see the teemi publication.