Antigenic Drift — 3D Strain Map

A random walk through hemagglutinin shape-space, mapped in 3D: horizontal drift over vertical time. Watch the vaccine committee's guess separate from where the virus actually goes.

Seasonal change magnitude5

Months: selection → flu season6

Surveillance sample quality70%

Prediction accuracy

—

Vaccine match

—

Expected effectiveness

—

Mismatch risk

—

Prior-season strain (start)

True circulating variant path

Committee's predicted / vaccine strain

Committee meeting point

Antigenic Drift as a 3D Random Walk Through Shape-Space

Every dot on the drift path is one month of real hemagglutinin mutation, walked as a correlated random step in a 3D antigenic map — the horizontal plane encodes antigenic distance, the vertical axis encodes time from the prior season to the next flu season.

Seasonal change magnitude: controls how large each monthly mutation step is, exactly like the 2D version's slider — but here it directly scales the length of a 3D random-walk vector.
Months: selection → flu season: sets how early the WHO-style committee has to commit to a strain, shortening how much real drift data they get to see before forecasting.
Surveillance sample quality: adds directional noise to the committee's read of the drift trend — lower quality means their forecast vector points further from the true trajectory.

What the geometry actually computes

The vaccine strain position is a genuine forward extrapolation of the (possibly noisy) trend vector measured at the meeting point; the final vaccine match is the inverse of the real Euclidean distance between that forecast and where the true random walk actually ends up twelve months in — not a lookup table, an emergent geometric result of the walk itself.