The SAMHD1 Function Atlas¶
Open the atlas → Best on a desktop browser. Drag to orbit, scroll to zoom, click anything.
Most descriptions of SAMHD1 are a list. This is the same material as a cell you can fly around inside: 54 functions placed in the compartment where each one happens, the links between them, and a switch that dims the whole board to what a single working copy of the gene can still do.
The embedded view above is fully interactive, but the atlas is designed for a full window — use the button for the real thing.
Four kinds of claim, not one list¶
The thing that makes SAMHD1 hard to describe is that its "functions" are not all the same kind of thing. Published lists put these side by side as equals:
- dNTPase activity — an enzyme doing chemistry
- Phosphorylation at T592 — something done to SAMHD1, by another enzyme
- Bladder dysfunction prevention — a whole-body outcome, several inferential steps away
They are not equals, and flattening them is what makes an otherwise accurate list feel untrustworthy. The atlas separates them into four tiers and lays the scene out along that axis, so distance on screen is distance in the argument:
| tier | what it is |
|---|---|
| Regulation | inputs that control SAMHD1 — the modification sites |
| Molecular function | what the protein itself does |
| Cellular consequence | what changes in the cell when a function is impaired |
| Body / clinical outcome | organism-level phenotypes, drawn outside the cell |
The canonical groupings¶
Reviews usually group SAMHD1's roles into four domains, and the rail lets you light up each one: dNTP balance (cell metabolism), genome integrity (tumour development), viral restriction (intrinsic immunity), and immune modulation (autoimmunity).
There is a fifth in this atlas — mitochondrial integrity — and it is marked as added here, because it is genuinely not in the standard figure. That framing predates most of the mitochondrial work. It is included because it is load-bearing for this variant: it is where the metabolic arm and the immune arm turn out to be the same arm.
Why half a gene is not half a function¶
The most useful thing the atlas shows is the WT / A565T switch, and the answer it gives is not the obvious one.
SAMHD1 only works as a tetramer — four copies assembled together. Someone with one variant copy does not make half as much working enzyme; they make tetramers of mixed composition. Any activity that needs all four subunits intact falls much further than gene dosage alone would predict, while activities that tolerate a mixed tetramer track dosage closely.
You can see that split directly: the dNTPase and the activities that run on it sit above 75% loss, while others land at 30–40%. That spread is the signature of a tetramer defect, and it is why this variant is not silent.
Two things are deliberately kept off that percentage scale, because they are not the same kind of measurement — the cancer associations are lifetime probabilistic risk rather than a per-cell deficit, and renal clear cell carcinoma is shown as protective, the one place where having less SAMHD1 helps. It is kept visible on purpose: an atlas that showed only harms would be an advocacy document rather than a model.
How confident is any of this?¶
Every node carries a grade, and so does every arrow:
- S — demonstrated for SAMHD1
- G — established elsewhere, imported here
- I — inferred; the reasoning, not a measurement
Two limits worth stating plainly rather than burying:
The nodes are sourced material; the arrows are our reasoning. The function list and its citations come from a literature compilation. The connections between functions are not in that source at all — they are authored for this atlas and each one carries its own grade. The "Where the evidence is thin" tour walks the weakest of them on purpose.
The location of each function is assigned, not quoted. No source we worked from records which compartment each activity happens in, and placing them is the entire point of a 3D map. The ring around each node grades where, not whether.
Twelve of the 48 citations are flagged where they are used — duplicates, references never resolved to a specific paper, preprints, a press release. They are shown rather than removed so a reader can judge them; verifying the remainder against PubMed is still outstanding work.
See also¶
The Immune Atlas picks up where this one leaves off: the mitochondrial arm here feeds the cGAS–STING loop there, and matching nodes in each link across to the other.
Start with the guided tours in the left rail — "What SAMHD1 actually does" is the five-minute version.