LAMP5, a surface antigen this project named as a fusion-enriched candidate, is 19.6x higher in normal cerebral cortex than in bone marrow - worse than every one of the 22 chromatin targets previously scored, including MEIS1
also called LAMP5 ratio 19.57 · LAMP5 cerebral cortex 41.1 nTPM · the worst normal-tissue ratio in this corpus
LAMP5 was on the shortlist of surface markers this leukaemia switches on. In a normal body it is about twenty times more abundant in brain cortex than in the bone marrow being treated, the highest such ratio measured here.
The same thing, for a scientist
`hyp-no-surface-antigen-is-both-fusion-driven-and-normal-tissue-clean-in-kmt2a-aff1` names **LAMP5** as one of its cases and carried no number for it. It now has one, on the same panel and metric as the 22 chromatin targets. **LAMP5: bone marrow 2.1 nTPM, highest normal tissue cerebral cortex at 41.1 nTPM, ratio 19.57** [E1]. For scale, the earlier run's worst of 22 was **MEIS1 at 17.2** and its most marrow-selective was CCNT1 at 0.4. **LAMP5 exceeds the previous worst case in this corpus.** 🔴 **And the signal is not a lymphoid artefact.** Its top two tissues are **cerebral cortex 41.1** and **ovary 38.0**, while lymph node is 5.3 and spleen 2.0, both near the marrow value. Unlike CD19 - whose 12.0 ratio is entirely spleen and lymph-node B cells - LAMP5's ratio is genuine parenchymal expression in brain, the compartment paediatric oncology least wants to dose. See `concept-the-leukocyte-artefact-inflates-a-haematopoietic-antigens-tissue-ratio-it-does-not-deflate-it`. 🔵 **This is the hypothesis's predicted shape, sharply.** A fusion-driven developmental surface antigen turning out to be a brain antigen is exactly the mechanism that entry proposes.
| script | analyses/hpa-normal-tissue-surface-antigens/run.py |
|---|---|
| status | run |
| run date | 2026-09-24 |
| gene | LAMP5 |
| ensembl | ENSG00000125869 |
| ensembl ID source | HPA-sourced (not independently verified) |
| bone marrow ntpm | 2.1 |
| highest normal tissue | cerebral cortex |
| highest normal ntpm | 41.1 |
| ratio worst to marrow | 19.57 |
top tissues — the numbers behind the chart
{
"cerebral cortex": 41.1,
"ovary": 38,
"pituitary gland": 6.6,
"lymph node": 5.3
}prior worst of 22
{
"MEIS1": 17.2
}measurements
[
{
"quantity": "highest normal tissue / bone marrow, HPA consensus RNA",
"value": 19.57,
"unit": "ratio of nTPM",
"subject": "LAMP5 (ENSG00000125869), cerebral cortex 41.1 nTPM over bone marrow 2.1 nTPM",
"universe": "a 16-tissue HPA consensus panel identical to the 22-target run; exceeds that run's worst case, MEIS1 at 17.2"
}
]- Neurotoxicity exposure - LAMP5 mRNA is about 19.6x higher in normal cerebral cortex than in bone marrow and the signal is parenchymal rather than an artefact of infiltrating leukocytes. Any LAMP5-directed agent needs a delivery argument - whether it crosses the blood-brain barrier - before the ratio can be dismissed. mRNA is not protein and this is not a toxicity prediction.
- Does any LAMP5-directed agent cross the blood-brain barrier? The ratio is only actionable if the cortex is reachable.
- Is LAMP5 protein present on normal cortical neurons at an engageable level? HPA immunohistochemistry would answer this and was not read.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.
In Human Protein Atlas consensus tissue RNA, LAMP5 is 2.1 nTPM in bone marrow and highest in cerebral cortex at 41.1 nTPM, a ratio of 19.57, with ovary second at 38.0 and the lymphoid organs lymph node 5.3 and spleen 2.0 near the marrow value.
What this evidence cannot show
Run 2026-09-24 (cycle 136) on the same 16-tissue panel, metric and identifier guard as analyses/hpa-normal-tissue-target-filter, whose two controls reproduced (ACTB 1.21 vs 1.2; ALB undefined in both). The Ensembl id was taken from HPA's own exact-symbol row and is HPA-sourced rather than independently verified. HPA consensus values are ADULT tissue and this is an infant disease. mRNA is not protein and expression is not dependence.
Evidence that argues against this (2)
- A high normal-tissue ratio is not disqualifying if the tissue is behind a barrier the agent cannot cross. Cerebral cortex is protected by the blood-brain barrier and a large antibody or CAR-T product may have little access to it, so LAMP5's 19.6 may be less actionable than a smaller ratio in an accessible organ such as PROM1's kidney at 5.27. This run ranks exposure by expression and takes no account of delivery.
- LAMP5's status as a fusion-enriched candidate comes from the surfaceome screen recorded in the hypothesis entry and was not re-verified here, so that premise rests on that entry's evidence rather than on this run.
LAMP5 at 19.57 in cerebral cortex exceeds the worst of the 22 chromatin targets
same run and panel; the parenchymal-versus-lymphoid reading applies to both
LAMP5 19.57 is unchanged; cerebral cortex is not a lymphoid organ