Each strand has a chemical directionality defined by its sugar-phosphate backbone: one end has a free 5′ phosphate group, the other a free 3′ hydroxyl. In the double helix the two strands run in opposite directions — one 5′→3′ upward, the other 5′→3′ downward. This antiparallel arrangement is required for the complementary bases to face inward and form hydrogen bonds at the correct geometry. You can visualise it here by noting that both backbone chains spiral in the same rotational sense but are offset by half a turn (π radians), placing each backbone node directly opposite its partner.

In the canonical B-form helix found in most cells under physiological conditions, there are approximately 10.5 base pairs per full 360° turn, giving a helical pitch of about 3.57 nm (10.5 × 0.34 nm). This number is not fixed: torsional stress introduced by topoisomerases or proteins can overwound or underwound DNA, changing the twist. In the simulator the Turns slider lets you freely explore different twist densities — try setting 60 base pairs and 6 turns to approximate the real B-form ratio, or reduce turns to model an underwound, open state.

The chemical structures of guanine and cytosine each present three complementary donor/acceptor sites when aligned in the Watson-Crick geometry: a donor-acceptor-donor pattern on one base perfectly matches an acceptor-donor-acceptor arrangement on the other, forming three hydrogen bonds. Adenine and thymine can only align two compatible sites. The extra bond makes G-C pairs roughly 30–40% stronger, so DNA regions rich in G-C require more energy — and higher temperature — to separate the strands, a property exploited in PCR primer design and genome stability analysis.

Clicking Randomise sequence generates a new random array of A, T, G, C bases for strand 1, then automatically assigns each position's complement (A↔T, G↔C) to strand 2, preserving strict base-pairing rules. The 3D geometry of the backbone is unchanged — only the colour-coded rung identities update. In "By base" colour mode you will immediately see the new distribution of red (A), green (T), blue (G), and yellow (C) rungs. Switching to Rainbow or Mono mode hides the base colours so you can study helical geometry without the sequence colouring.

The simulator uses the real biochemical value of 0.34 nm per base pair to calculate the "Approx length" figure in the stats panel, so for 30 base pairs it correctly reports 10.2 nm — consistent with crystallographic data for B-form DNA. The 3D rendering is scaled up for visibility (approximately 3× the physical size), so the on-screen model is not proportional to the actual molecule; the stat readout is. The helix radius in real B-DNA is about 1 nm, so the default slider value of 1.0 provides a reasonable starting approximation of the true diameter.