Genetics and Genomics DNA Helix Explorer (2D): Variant to Protein, Primer Tm and Hardy-Weinberg

Edit a short gene as a ladder of base pairs: see how each change alters the codon and the protein, how a PCR primer melts, and how common the carriers are.

Switch to 3D

Presets

Playback

Parameters

Model check

The science

From DNA to protein. A gene's coding strand is read three bases at a time. The standard genetic code maps the 64 possible codons to 20 amino acids and 3 stop signals, so most amino acids have several codons (leucine, serine and arginine have six). In a cell the ribosome reads messenger RNA, which carries the coding-strand sequence with U in place of T; the model therefore translates the coding strand directly. The toy gene on the helix is a short synthetic sequence invented for this page, not a real human gene.

Variant effects. A one-base change inside a codon is called silent if the amino acid stays the same, missense if it changes, and nonsense if it creates a stop. Missense changes are split into conservative and radical with a simple rule: radical if the charge class changes (positive, negative or neutral) or the Kyte-Doolittle hydropathy changes by 3 or more. This is a teaching heuristic, not a clinical predictor: real interpretation also needs protein structure, conservation and experiments. Deleting or inserting one base shifts the reading frame for every later codon (frameshift). Losing the start codon ATG or the stop codon TGA is also shown. The preset "Sickle-type: E to V" is the same kind of change (a glutamate replaced by valine) that underlies haemoglobin S, which in real HBB occurs at codon 6; here it is at codon 5 of the toy gene.

Counting all changes. Of the 61 × 9 = 549 possible one-base changes of sense codons, 134 are silent, 392 missense and 23 nonsense with the standard code, which is why roughly a quarter of random point mutations do not change the protein. The page counts them by enumeration, and cross-checks the nonsense count from the stop-codon side.

Primer melting. A PCR primer sticks to its complementary strand until the temperature reaches its melting temperature Tm, the point where half of the strands are paired. The nearest-neighbour model sums the enthalpy and entropy of each stacked base pair (SantaLucia 1998 parameters), adds end and symmetry terms and a salt correction ΔS(Na) = ΔS + 0.368 (N−1) ln[Na+], then Tm = ΔH / (ΔS + R ln(CT/4)) with CT the total strand concentration. The quick rules shown beside it are Wallace (2 degC per A or T, 4 degC per G or C, meant for short oligos under about 14 nt), Marmur-Doty (64.9 + 41 (GC − 16.4)/N) and an empirical salt-adjusted formula (81.5 + 16.6 log10[Na+] + 0.41 %GC − 675/N). They are approximations and can differ from each other by several degrees.

Hardy-Weinberg. For a variant with allele frequency q, random mating in a large population gives genotype frequencies p² + 2pq + q² = 1 with p = 1 − q. A rare variant (q = 0.005) has about 1 carrier in 100 people but only 1 homozygote in 40 000. The law assumes no selection, migration or mutation drift and random mating, so real populations deviate from it.

Honest simplifications. Only a 12 to 48 bp stretch is drawn, as a flat ladder of base pairs (no twist, bends, nucleosomes or supercoiling). Introns, splicing, promoters, codon usage and tRNA are not modelled, the amino-acid beads are schematic and drawn above their codons. Insertions and deletions are simple one-base edits. Kyte-Doolittle values are the published scale, and the nearest-neighbour numbers are quoted from the literature at 1 M NaCl then salt-corrected.

Checks. (1) the genetic code table: counts of codons, stops, amino acids, degeneracy and the nonsense count from both sides; (2) Hardy-Weinberg against a 200 000-person random-mating simulation (chi-square) and allele-frequency recovery; (3) the Wallace rule for a 12-mer against its hand value, and the nearest-neighbour Tm against a bisection of the mass-action duplex fraction; (4) the translation of the reference gene against a hand translation, double reverse complement, equal GC on both strands, the variant classifier on six known changes . See also the pages on the helix and its melting and the replication fork, which cover physical melting and copying.

Pick a preset variant, or click a base and click it again to change it. Red flags mark changed bases.
Hardy-Weinberg genotype frequencies against allele frequency q
Kyte-Doolittle hydropathy of the reference protein and of this variant
Primer melting temperature against primer length (four methods)