// research/notebook · v0.42 · seed d9894b85
A developer's lab notebook exploring chromosomal biology through generative code. Telomeres shorten, replication forks split, base pairs hum a four-letter song — we render the song.
Adjust the parameters. Watch the chromosome cap shrink with each cell division.
// each division strips ~50 bp from the cap
function divide(chrom: Chromosome): Chromosome {
const loss = 50;
const rate = 1.00;
const gens = 40;
return {
...chrom,
cap: Math.max(
0,
chrom.cap - (loss * rate)
),
age: chrom.age + 1,
senescent: chrom.cap < 2000
};
}
// simulate 40 generations
let chrom = { cap: 10000, age: 0 };
for (let i = 0; i < gens; i++) {
chrom = divide(chrom);
}
// → cap: 8000 bp · senescent: false
Field notes from the boundary between code and biology.
Treated each nucleosome as a node and Hi-C contact frequency
as edge weight. The 3-D embedding self-organizes into the
canonical A/B compartment pattern within ~600 iterations.
Code splits cleanly into load.ts,
solve.ts, and render.glsl.
Replotted Blasco et al. data with the four-color base-pair palette below. The TERT-positive cohort shows a clear plateau where the mortal cohort decays linearly. The figure ships as an SVG sparkline embedded in the post.
Wrote a zero-dependency parser for SAM/BAM CIGAR operations.
Streams over the input, allocates nothing, handles the long
N skips found in spliced RNA-seq alignments.
Benchmarks at 1.2 GB/s on the M-series test box.
Judson's account of Crick walking into the Eagle still hits differently after a decade in software. There is a kind of debugging in molecular biology that programmers immediately recognize — print statements, but the printer is a gel.
A randomly-walking strand. A · T · G · C, scrolling at the speed of replication.
// fork · clone · branch
Open notebooks, open data, open source. The repository
is mirrored at git://telomere.dev/lab.