This project explores absent and shared DNA words in the included bacterial genome sequences. The main workflow is the documented nullomer and shared-sequence notebook.
The repository includes a static, browser-based app for exploring sequences:
open the Genome Words app.
The Deploy static app to GitHub Pages workflow builds and deploys it on pushes
to main and can also be run manually from GitHub Actions. If Pages has not
been enabled for this repository yet, set Settings → Pages → Build and
deployment → Source to GitHub Actions.
The app can load the included E. coli K-12 and C. metallidurans CH34 example genomes, or users can paste or upload their own sequence files. It runs locally in the browser and includes an optional link to search a shared sequence with NCBI BLASTN. The absent-word search is capped at 10 bp to keep exhaustive enumeration practical in a browser.
For a local preview, run python -m http.server 8000 in the repository folder,
then open http://localhost:8000.
- Open
god_sequence.ipynbin Jupyter or VS Code. - Select a Python 3 kernel.
- Run all cells from the repository root so paths under
data/resolve.
The notebook uses only the Python standard library. It analyzes the sequence
files already included in data/ (headerless sequences despite their .fasta
suffixes, also compatible with ordinary FASTA records); no downloads or package
installation are needed. Change GENOME_FILES in the notebook to select other
genomes. The longest shared sequence is determined from the input lengths; it
is not limited to a preset k-mer size.
The notebook also includes an optional NCBI BLASTN step for a selected longest
shared sequence. It is disabled by default. Enabling it sends the sequence to
NCBI's public nt database through the
BLAST Common URL API.
Review that disclosure before setting RUN_NCBI_BLASTN = True. The cell waits
for the submitted search and prints a link to its results.
- A k-mer is a contiguous DNA word of length k.
- The notebook reports all shortest absent words in E. coli K-12, searching
from length 1 up to
MAX_NULLER_LENGTH. A result at the configured limit is bounded: it does not establish that no longer absent words exist. - The notebook finds the longest shared sequence: the longest exact, contiguous DNA substring present in every selected genome. Only maximal matches are returned, so their shorter substrings are omitted. The result is for the selected input files, not a universal claim. This requested shared-sequence analysis is distinct from the paper's definition of primes as the shortest sequences not found in any reported genome.
- FASTA records and separate input files stay separate, and ambiguous symbols
such as
Nbreak sequence runs. This prevents matches from crossing contig boundaries or unknown bases. Reverse complements are treated as distinct words. - Absence alone does not establish biological function, lethality, or a reason for absence. The notebook does not enumerate every minimal absent word (MAW).
Hampikian and Andersen, 2007, Absent Sequences: Nullomers and Primes, Pacific Symposium on Biocomputing 12:355–366.