A curated, evidence-graded database of mTOR pathway research -- 322 studies and 120 cross-linked entities (45 with their own page), each claim traced to a primary source and rated by strength of evidence. This page explains who curates it, how a study earns a place, what the grading does and doesn't guarantee, and how to report an error.
Oliver's mTOR Atlas is a narrow, hand-curated corpus -- not an attempt to index all of PubMed's roughly sixty thousand mTOR-related records, but a smaller set small enough that every entry can be read, graded and defended by one person, then connected by hand into a knowledge graph of genes, drugs, diseases and outcomes. Every claim carries an explicit evidence tier (A = systematic review of human data, B = human trial, C = animal model, D = mechanistic/in-vitro/review), and the corpus deliberately keeps negative results -- studies where a popular longevity compound did not extend lifespan -- with the same visibility as positive findings.
Oliver Barton -- Creator & Curator, Prague, Czech Republic, age 15. A high school student with a self-directed research interest in mTOR signaling and evidence-based science curation, who built the Atlas to be the structured resource he wished existed when he started reading primary literature on the pathway: studies, mechanisms and honest evidence grades held side by side, rather than scattered across review articles.
There is no editorial board and no second reviewer -- see "Who reviews the selection" below for what that does and doesn't mean for trust.
Contact: oliver.barton1113(at)gmail.com. Corrections, questions and error reports are welcome, and are logged in the project's GitHub commit history, which is public.
Every study passes through the same four steps before it's added:
Every DOI resolves to the publisher's own page (Nature, Cell, Science, NEJM, Lancet and others), so any claim traces back to the original paper in one click.
A study earns a place if it does at least one of four things: establishes a landmark mechanism in mTOR biology; supplies the strongest available human evidence for a claim (a systematic review or a large RCT, not an isolated small trial); fills in representative animal or mechanistic work for a pathway node that would otherwise be uncovered; or reports a negative or null result that the rest of the literature tends to under-report -- included deliberately, since a database that only shows what worked misrepresents the actual state of the science.
A candidate is left out if it's a secondary commentary, editorial or news piece rather than a primary study or review; if it duplicates a pathway node already covered by a stronger study and adds no new claim; if its metadata can't be verified against PubMed (no PMID or DOI resolving to the publisher's own record); or if its connection to mTOR is incidental -- mTOR measured as one readout among many in a paper about something else. Preprints and registered trials are admitted only when they're the best available evidence for a claim, and are then labelled ungraded rather than given an A–D tier.
An honest limitation: the Atlas does not keep a screening log. Candidates that were considered and rejected leave no record, so no exclusion count can be quoted or inferred. That's a real weakness compared with a systematic review, where the screening flow is itself evidence that the search was unbiased -- here it isn't. Selection is a judgement call, made paper by paper, and it is not currently auditable from the outside. The Atlas should be read as a curated reading list with grades attached, not as a systematic review of the mTOR literature. Logging rejected candidates and the reason is a planned change.
Who reviews the selection: one person, the curator. There is no second reviewer or independent adjudication of borderline calls -- the usual safeguard against a single reader's blind spots is absent. The one external check to date was an unsolicited scientific review in July 2026, which raised sixteen points; all were addressed rather than quietly dropped, and one exposed a real inconsistency between the stated A–D tier definition and how it was actually being applied. An automated validator rule now blocks any deploy where a study's tier and its underlying evidence level disagree.
An automated job screens PubMed for new candidate mTOR studies and relevant conferences once daily, at 02:00; the decision to include and grade a flagged paper is made by hand, and publishing the updated site is a manual step. So the screening is daily, but the corpus changes only when a human accepts a candidate -- typically a handful of papers a month, sometimes none. The exact timestamp of the live corpus is printed in the footer of every page, and every count on this site, including the ones above, is computed from that snapshot rather than typed in by hand.
Content is CC BY 4.0 -- free to cite and reuse with attribution to "Oliver's mTOR Atlas". The dataset is archived and citable via Zenodo, concept DOI 10.5281/zenodo.22059963.
