Each VictorOS map shows a single rare disease as a landscape. The disease sits at the center, and everything known about it radiates outward: the genes and proteins beneath it, the biological mechanisms it sets in motion, the symptoms and findings people actually experience, and the treatments and tests that act on it. The picture carries a lot of information at once. This page explains the visual grammar so you can read any map with confidence.
A map is made of nodes (the circles) and links (the lines between them). A node is one thing the research talks about: a gene, a protein, a biological pathway, a symptom, a treatment. A link is a relationship the research establishes between two of them, for example a gene that drives a finding, or a drug that treats a symptom.
The large circle near the center is the disease itself. The nodes cluster around it according to how the layout settles rather than by importance: nodes pull toward whatever they connect to and push away from everything else, so densely connected ideas gather near the middle and looser ones drift to the edge. You can drag any node to move it, click a node to open its full provenance in the side panel, and use the layer checkboxes to hide or show parts of the map. Nothing you do changes the underlying data. You are only changing your view of it.
Four visual channels carry the meaning: the layer a node belongs to, its color, its size and label, and the style of the lines. The rest of this page takes them one at a time.
Every node and link belongs to one of five layers. The layer says how firmly the claim is established, which is the single most important thing to read on the map. The checkboxes in the top-left of the map turn each layer on and off.
The confirmed core. A claim that a named, accountable reviewer has checked against its source and entered into the permanent record of the disease. Canon nodes are drawn solid, slightly larger, and always carry a label. This is the most trustworthy layer.
Sourced, awaiting promotion. Something drawn from a published source and recorded, still waiting to be elevated into canon. Observation nodes are drawn solid too. Most of what you see on a newer map lives here.
A proposed connection. A relationship that has been suggested but not confirmed. Drawn with a dashed, fainter line to signal that it is not settled.
An early hint. A pattern that has surfaced in the data and may be worth attention, but stands on thin or preliminary evidence. Dashed and faint.
An open question. Something the map deliberately marks as unknown or unresolved rather than leaving blank. Dashed and faint.
The shorthand to remember: solid means established (canon and observation), and dashed means unproven (hypothesis, signal, question). Unproven nodes are also drawn slightly more transparent. If you want to see only what is firmly grounded, untick the bottom three boxes.
Color encodes the type of each node, what category of thing it is. The legend on the map shows only the types that actually appear in that particular disease, so no two maps show the same set. Here is the full palette.
You do not need to memorize these. The point is that related kinds of thing share a family of color, so the genetics of a disease, its mechanism, its symptoms, and its treatments each read as a region of the landscape.
This is the question the map raises most often. The short answer: a node grows with the amount of evidence behind it, and only nodes past a size threshold are labeled permanently, to keep a crowded map readable.
A node's size reflects how many separate observations support it. A node mentioned once is small. A node confirmed across many sources is large. Canon nodes get a small size bonus on top, because confirmed claims earn prominence. The growth is capped, so a heavily supported node does not balloon off the screen.
If none of these is true, the label is hidden to reduce clutter, but it is not gone. Hover over any small node and its label appears. So the unlabeled nodes around the edge are not hidden information; they are the more thinly supported findings, with their names a mouse-over away.
One caveat worth stating. This rule ties a node's visual prominence to how often it is cited rather than to how clinically important it is. A finding that matters a great deal but appears in only one or two sources will look small. Size reflects evidence weight, so when a node is small, treat it as an invitation to click and read what stands behind it.
The lines carry their own grammar. The style of a line tells you both how settled the relationship is and where it came from.
A relationship in the canon or observation layer. Drawn as a solid line. This is a connection the record stands behind.
A relationship in the hypothesis, signal, or question layer. Drawn dashed and faint, to match its unproven nodes. Proposed, awaiting confirmation.
A relationship where serious sources disagree. Drawn in red and slightly heavier. The disagreement is kept visible rather than resolved away.
There is also a fourth style: a faint purple line marks a derived link, one the system inferred by reasoning over other relationships rather than reading it directly from a single source. Click either node a line connects to see the relationship spelled out in the side panel, with its own source.
The map is the overview. The panel on the right is where you check the evidence. It has three tabs.
Click a node and this tab shows its type, layer, evidence rating, observation count, a plain-language summary where one exists, the source it rests on, and every relationship into and out of it. Source markers like a PMID link straight to the study on PubMed.
A set of standing questions answered from the graph, such as approved drugs that act on the disease's pathways, recognized subtypes, and any symptoms that do not yet trace to a mechanism.
Where the map flags internal tensions or claims due for re-examination as the science moves. Some checks are automatic; the deeper ones remain a human review.
The header above the map always shows the running totals for that disease: how many nodes and links it holds, how many are canon, and how many links are disputed. That single line is a quick read on how mature and how contested a given map is. To follow a single claim all the way back to its source, see how the method works and how to verify it yourself.
The grammar makes sense fastest with a real disease in front of you. Pick a guide, switch to its map, and try clicking a large node and then a small one.
VictorOS organizes evidence; it does not practice medicine. It does not give medical advice, diagnose, or recommend treatment. It builds the evidence landscape so patients, caregivers, and advocates can engage their care teams as informed participants, bringing sourced evidence into the conversation. These guides support your medical team; your clinicians remain the ones who diagnose and treat.