From a TP53 variant to an evidence-anchored hypothesis in one session.
Enter any variant and read what is actually measured: the Measured binding card searches the public ChEMBL bioactivity archive live for your specific variant, plus same-residue context, and surfaces published Kd, Ki, IC50 and EC50 values with their assay method and original citation. Cohort frequency, functional screens and cited literature sit alongside. The computational shortlist is a transparent physics screen for cavity fit and ligand efficiency, anchored in measured data and published evidence — docking scores are never presented as binding affinities. Every number auditable, every export reproducible.
More than half of human tumours carry a TP53 mutation. Understanding what each variant breaks, and what could reactivate it, is one of oncology's highest-stakes questions.
HyperLab53 folds the steps a cancer biologist would otherwise chase across many tools, variant entry, receptor selection, mutation-aware modelling, WT overlay, cavity map, live measured-binding lookup against the public ChEMBL archive for your specific variant plus same-residue context, ligand screening against 3,311 approved drugs and a curated rescuer library, per-residue interaction analysis, evidence-linked rescue shortlisting, orthogonal in-cell validation, and a reproducible screen export, into a single continuous session across roughly 4,000 TP53 variant profiles.
The problem, the gap, and what HyperLab53 actually does about it.
TP53 is mutated in over half of human tumours, yet direct reactivators remain rare and mutation-specific. HyperLab53 collapses the mutation → structure → pocket → ranked ligand loop into a single transparent session across roughly 4,000 TP53 variant profiles and 3,311 approved drugs, and shows its work at every step.
One mutation, one target, years of setup.
Each p53 hotspot (R175H, R248Q, R273H, Y220C, R282W and beyond) has a distinct destabilisation and pocket signature. Standard pipelines rebuild the structural and screening context from scratch per variant.
Pathogenicity scores stop short of a rescue hypothesis.
General pathogenicity scores classify a variant as damaging but do not identify the disrupted structural motif or the ligand class most likely to reactivate it. HyperLab53 targets that mechanistic gap.
Mutation-aware, mechanistic, and honest.
Mutation-aware candidate model with WT overlay and cavity map, deterministic physics-first screening against 3,311 approved drugs and a curated rescuer library (cavity-fit prefilter, then docking, then ligand-efficiency ranking), curated motif attribution, calibrated uncertainty bands, and DMS overlays, all exported as JSON across roughly 4,000 TP53 variant profiles.
From mutation to rescue candidate
Eleven stages, one continuous workflow, sequenced so a cancer biologist can move from a variant to an evidence-linked candidate without leaving the session.
Variant entry
Enter a TP53 protein change or upload a catalogue covering up to roughly 4,000 variants. The workspace keeps only variants that can be structurally folded and docked, so the rest of the session works on a well-defined set.
Receptor selection
Choose an experimental PDB, upload a private PDB or CIF, or use a mutation-aware candidate model when no experimental structure fits the variant.
Interactive 3D view
The 3D viewer keeps working when optional volume servers are unreachable. Click a residue in the sequence to focus on it in the structure with legible atom, residue, and element labels at any zoom level.
WT overlay and cavity map
Kabsch RMSD, per-residue displacement, and a geometric grid-scan pocket finder (including metal cofactors) surface cavities around the mutation site. Static model. Dynamics-opened cryptic status still needs MD.
Ligand screening, de-novo generation and hand-drawn design
Screening runs in three lanes with separate session records. Screen Library: Stage 1 ranks the whole library (381 curated rescuers plus, optionally, 3,311 approved drugs) by physicochemical cavity fit using Chem Toolkit descriptors; Stage 2 docks only the best-fitting candidate with a Lennard-Jones plus Coulomb plus H-bond force field. Generate De Novo: a fragment-masked generator proposes pocket-sized structures that are filtered to the cavity's heavy-atom window and docked on the same terms. Design & Dock: you draw a molecule yourself in the built-in structure editor; it is parsed, sanitised and size-checked by the Chem Toolkit, saved per variant in your own browser, and docked through the identical pipeline - tagged 'manual design - no literature evidence' and never mixed into the screening or de-novo counters. Each lane runs one compound per click, with explicit add-more increments of 1, 5, 10, 25, or 50. Receptor can be a catalogue PDB, a candidate model, or an uploaded structure.
Top-N docking and comparison
Docked hits are ranked by ligand efficiency (kcal/mol per heavy atom) so large scaffolds do not win on size alone, then reloaded into the 3D viewer. Every pose passes a structure-validation gate (parseable MOL block, resolvable elements, valid bond orders, genuine 3D spread, no NaNs) before it is drawn, and the viewer shows one compound at a time by default with a configurable in-scene cap. A side-by-side dialog compares the docked set with Chem Toolkit descriptors on one screen. Literature evidence is shown as a label, not a filter.
Per-pose diligence
Every dock is followed by a geometric ligand-residue interaction scan (H-bond, hydrophobic) and a Pose Checks-style geometric sanity check (clashes, bond lengths, containment, compactness, element sanity).
Motif attribution
For curated hotspot variants, static SAE-derived profiles map the mutation to disrupted structural motifs (DBD hydrophobic core, L2/L3 zinc coordination, DNA-contact loop) alongside per-residue ΔΔG.
Experimental overlays
Deep-mutational-scanning data, the Sánchez-Rivera Lab prime editing sensor screen, and a live ChEMBL measured-binding lookup are joined per variant so published in-cell fitness and real assay affinities sit next to the prediction.
Rescue-vs-inhibit reading
Every candidate is scored as rescuer versus inhibitor with transparent features and a calibrated uncertainty band. RescueBench splits its evaluation into the Rescue task (curated compound-mediated reactivation triples - what the scorer is designed for) and the LoF-tolerance task (MaveDB Nutlin/Etoposide fitness - variant tolerance, not rescue), and reports AUROC separately for hotspots vs. rare mutations plus a leave-one-mutation-out row that isolates structural inference from memorized literature. Every per-mutation Rescue Brief exposes a deterministic shortlist gate - panel-active (≥0.40), matched-class, or broad-class+lit - and RescueBench reports Precision@10, Precision@25 and Recall@shortlist on that same gate so precision and recall are directly comparable to the UI.
Structured screen export
The full visible screen exports as a single organised JSON snapshot: mutation metadata, active structure with provenance, every card value, tables, links, docked poses, per-pose diligence, and citations. Fields that have not been computed in this session are null, never fabricated.
Continuous workflow
Variant entry, receptor selection, mutation-aware modelling, docking, per-pose diligence, and evidence-linked shortlisting live in one screen so the biologist keeps context between steps.
Interactive 3D residue view
Click a residue in the sequence and the 3D viewer focuses on it with legible atom, residue, and element labels that stay readable at any zoom. Rapid consecutive clicks cancel prior transitions without jitter.
In-browser chemical space
Every library, docked and generated compound is fingerprinted with ECFP4 (Morgan r=2, 2048 bit) in a background worker and cached in IndexedDB. Tanimoto similarity search, a descriptor-based scaffold-hop view, exact Chem Toolkit descriptors with PAINS/Brenk alerts and a documented 0-100 developability triage index, a 2D PCA map of the filtered set, high-confidence mode, browser-local annotations, and CSV export of the filtered or selected set. Descriptor arithmetic only - no machine-learned ADMET numbers.
Side-by-side ligand comparison
Compare the docked shortlist in one dialog: 2D structures with Chem Toolkit-derived MW, cLogP, HBD/HBA, TPSA, rotatable bonds, and Ro5 violations arranged so the full set is legible in a single viewport.
Ligand ↔ residue interactions
Every dock is followed by a deterministic geometric scan for hydrogen bonds and hydrophobic contacts between the pose and each receptor residue.
Pose sanity checks (Pose Checks-style)
Core Pose Checks (Buttenschoen 2024) geometric checks (clashes, bond lengths, containment, compactness, element sanity) run against every pose alongside the score.
Modular generative screening pipeline
Generation and evaluation are kept strictly separate. The cavity finder measures the pocket, its volume is converted into a heavy-atom budget, and that budget is sent as an explicit size constraint to the de-novo generator and the lead optimiser - so oversized molecules are never proposed. Returned structures are flat SMILES only: they are re-parsed with Chem Toolkit, re-filtered to the same heavy-atom window, embedded in 3D by distance geometry, and docked with the same in-browser physics engine as every approved drug. Binding energy, pose validity and ranking always come from local physics, never from the generative model, and generated molecules are excluded from every benchmark and validation total.
Physics-first rescue shortlist
The docking table ranks the library by cavity fit before any compound is docked. Known rescuers are labelled as known, but novel hits can reach the table purely on geometric and physicochemical fit, so the tool proposes pocket binders rather than only reciting literature.
Approved-drug repositioning sweep
The 3,311 approved drugs in the ChEMBL max_phase 4 library can be screened against any TP53 variant pocket in the same session. Stage 1 ranks them by cavity fit; Stage 2 docks the best-fitting batch, so repositioning ideas surface next to mechanism-class rescuers, not in a separate tool.
Live measured binding lookup
For any variant you enter, the Measured binding card queries the public ChEMBL bioactivity archive live, plus same-residue context, and surfaces published Kd, Ki, IC50 and EC50 values with assay method, construct and original citation. No measurement found means exactly that - nothing is imputed.
Live disease and homology context
Top TP53 disease associations from the Open Targets Platform and structurally similar PDB entries from the RCSB Search API sit next to the loaded receptor. Bioactivity queries have a secondary source, so a single upstream outage does not stall the session.
Pharmacophore and hydrophobicity
Chem Toolkit-JS identifies donor, acceptor, aromatic, hydrophobic, and ionizable groups on the ligand; a Kyte-Doolittle heatmap around the mutation position summarises local hydrophobicity from the loaded PDB.
Retrosynthesis disconnection hints
Chem Toolkit-JS scans the docked ligand for named-reaction retrons (amide, ester, sulfonamide, urea, carbamate, Suzuki biaryl, Buchwald C to N, aryl and Williamson ethers, reductive amination, Wittig, click triazole) and lists every bond a known coupling could plausibly have formed. Single-step retrons only; not a full route search.
Live NCI GDC cohort context
For the selected TP53 variant, the lab pulls live from the NCI Genomic Data Commons (api.gdc.cancer.gov): per-project case counts and cohort frequencies across TCGA, TARGET, CPTAC, and MMRF, plus the top co-occurring mutated genes in the same carrier cases. Overlay only - scores and rankings are unchanged; the SSM UUID links straight to the GDC portal.
Mutation evidence explorer
A structural / functional / clinical summary card gathers the mutation's pocket class, ΔΔG, DMS and prime-editing sensor readouts, GDC cohort frequency, and curated literature into a single collapsible view so context sits next to the prediction.
Residue overlay track
A 1D sequence track aligns per-residue DMS scores and prime-editing sensor values along the p53 canonical sequence, aggregated as the maximum score per codon so hotspot signal is not diluted by silent variants.
Rescue-vs-inhibit scatter
On every candidate page a 2D scatter plots pRescue against pInhibit for the docked shortlist, so rescuers, inhibitors, and ambiguous compounds are separated visually before any ranking decision.
Europe PMC evidence backbone
Every step of the workflow - variant, receptor, pocket, each docked or drawn ligand - is checked live against Europe PMC's full-text life-science index. The Evidence backbone card returns a per-step verdict with the papers behind it, so a recommendation either has published support attached or is explicitly marked as unsupported.
Batch mode and run history
Queue several variants in Batch mode and re-open any previous session from Run history: every dock, generated structure and hand-drawn design is written to browser-local storage per variant, so signing out and back in on the same browser restores the shortlist, poses and notes exactly as left.
Re-measurable benchmark ledger
Methods and Benchmarks pages do not just quote numbers - screening enrichment, affinity calibration, pose site labelling, cavity detection and pose accuracy can each be re-measured in your own browser on demand, and the ledger records the value your machine produced next to the published one. Literature evidence can be re-queried on demand from any run, so a long batch never ranks on stale citations.
Manual design lane (structure editor)
Draw any molecule in the built-in Ketcher editor, save it per variant, and dock it through the identical physics pipeline. Manual designs carry a 'no literature evidence' tag, stay in your browser, appear in the JSON export, and are never mixed into screening, de-novo or benchmark counters.
A-E candidate grade on every hit
Each docked compound carries a single letter grade that folds pose quality, site relevance, geometric sanity, ligand efficiency and - where it exists - published evidence into one readable verdict. Clicking the grade opens the full breakdown with every contributing factor, its weight, and the receptor's provenance (experimental crystal versus predicted model), so a biologist can triage a shortlist without reading four numeric columns.
Literature-anchored ranking
Ranking is not physics alone. Where published work exists for a compound against the exact variant or its mechanism class, that evidence is scored and blended with the pose result, so clinically-relevant rescuers surface at the top instead of being buried under geometrically lucky, unpublished molecules. Every evidence contribution is itemised and traceable to the papers behind it.
De-novo protein binder design
Beyond small molecules: generate a mini-protein backbone aimed at the mutant pocket, design a sequence onto it at a binder length you choose, request several candidates, and load any of them into the same 3D viewer as an extra layer beside the receptor to see exactly where it engages the mutation. Several sequences are screened by a fast foldability pre-check before any folding job is spent, and every binder carries a buried-surface-area affinity estimate with its uncertainty alongside the geometric acceptance criteria.
Honest statistics on small screens
A screen of a handful of molecules no longer returns a blank. Runs of roughly three to seven compounds report an enrichment factor with a 95% confidence interval resampled from the run itself, so an interval straddling 1x says plainly that the ordering is not yet distinguishable from chance; the stricter ranking statistics still require eight ranked compounds and two with prior evidence.
On-demand stability prediction
For any plain single-residue substitution, the evidence explorer can call an independent published structure-based ΔΔG predictor on the p53 crystal and show its value next to the measured literature entry and the local model. It is opt-in, a measured value always wins, and an unreachable service is reported as unreachable rather than filled in.
Cavity detection, rescoring and receptor QC
Pockets are detected with an in-browser port of the P2Rank residue-scoring approach plus a geometric cavity fallback, cached locally so no session stalls on an external server. Each candidate pocket can be rescored against the mutation neighbourhood, and a receptor completeness gate flags missing backbone or side-chain atoms and chain breaks before any pose is trusted.
Consensus and ensemble docking
A pose does not have to rest on one run. Multiple receptor conformations and independent scoring passes are compared, agreement between them is reported, and rank-fusion combines the signals - so a hit that only wins under one particular setup is visible as such rather than presented as a robust result.
Suggest the next experiment
For the current variant and shortlist, the lab proposes the concrete bench step that would most efficiently discriminate the hypothesis - which assay, which construct, which control - and records the expected outcome, turning a computational shortlist into a testable plan.
Batch upload from CSV, TSV or Excel
Drop a spreadsheet of variants and the batch runner queues each one through the identical pipeline, writing results to run history so a whole panel can be triaged in one sitting rather than variant by variant.
Reproducible screen export
One click exports the visible screen as organised JSON: mutation metadata, active structure with provenance stamp, per-mutation Rescue Brief (structured payload plus Markdown), mutation-intel enrichment, prime-editing sensor scores, external affinity + MMGBSA benchmark blocks, live GDC cohort and co-mutation context, Europe PMC evidence verdicts, chemical-space index state, manual designs, and a letter-for-letter verbatim snapshot of every card on screen with its machine-readable payload. The export also carries the small-screen enrichment interval, the binder foldability pre-check and buried-surface affinity estimate, and any on-demand stability prediction. A second option packages the same data as a full bundle - Markdown report, CSV tables, figures, per-card screenshots, and the pose plus receptor structure files - and both are stamped with a content hash so identical screens produce identical exports. Uncomputed fields stay null, never fabricated.
Powered by measured analogues, not a black box
Docking scores are never converted into an affinity. The only affinity value shown beside a candidate is read across from the nearest measured analogues in a 4,500-compound set of published affinities (HiQBind, BindingMOAD, GatorAffinity p53 records), as a similarity-weighted mean over the three closest molecules above an ECFP4 Tanimoto cutoff, and it is labelled as read-across wherever it appears. On held-out validation the close-analogue tier reaches mean absolute error 1.14 pKd with 80% within ±1.80 pKd, and every value carries an honest conformal ±band. A ligand-only neural predictor was benchmarked on the same held-out set and failed it (Pearson r = -0.23), so it is not used for the reported column.
4,500 affinities
Read-across draws nearest neighbours from published measured affinities (HiQBind, BindingMOAD, GatorAffinity p53 records), not from a model's training set.
MAE ≈ 1.14 pKd
Held-out validation on 800 measured complexes. 80% within ±1.80 pKd at the close-analogue tier (Tanimoto ≥ 0.5, 52% coverage). Honest conformal ±bands, not an extrapolation.
No upload, no queue
The ECFP4 similarity search and read-across run client-side — every screened or generated molecule is scored straight from its SMILES string, with nothing sent to a server.
Real numbers, cited sources, one continuous workflow.
Total experimental and computational measurements across 8 public benchmarks (1,300,000+ unique complex-with-affinity records + 374,518 pose-RMSD measurements; per-source counts on /rescuebench)
Protein-ligand complexes in the pose-quality set (BindingNet High MCS RMSD); median heavy-atom RMSD 0.77 Å, 86.7% below 2 Å
Pathogenicity concordance on the 10-variant primary validation set (all 10 variants predicted pathogenic)
Rescue-ligand classification F1 on the 10-variant primary validation set (PASS vs non-PASS; 6 TP, 2 FP, 2 TN, 0 FN)
Pathogenicity concordance on the 51-variant curated panel (47/51 variants correctly classified)
Approved drugs (ChEMBL max_phase 4) available for in-browser repositioning docks against any variant
TP53 variants searched live for measured binding data in ChEMBL; most return no published affinity, and the card reports that honestly rather than fabricating a number
Curated TP53, MDM2, and MDM4 rescue ligands sourced from DGIdb, ChEMBL, and PubChem
Full-text life-science articles indexed in Europe PMC, queried live to anchor every pipeline recommendation in published evidence rather than opaque scoring
Reachable TP53 DBD missense variants (4,137 missense in the catalogue; the dockable subset used for structural inference)
Deterministic physics core
Structure viewing, ligand geometry, docking, and scoring run through a transparent Lennard-Jones plus Coulomb plus H-bond force field with simulated annealing. No opaque model file, no hidden calibration.
Prediction, on demand
When a mutation has no suitable experimental structure, a mutation-aware model is generated and streamed back to the viewer across roughly 4,000 reachable TP53 variant profiles.
Provenance you can filter
Every card carries a provenance tag - measured, measured-modelled or derived - and a global toggle can withhold everything that is not a wet-lab measurement. Dataset row counts, molecule coverage and validation checks are published on the data-quality page.
What the workspace helps a p53 biologist do
A single screen carries a p53 variant from selection to a scored, evidence-linked shortlist. The per-mutation view sits next to experimental data and citations, and every visible result is exportable as a single reproducible snapshot.
- Mutation-aware candidate structures when no experimental PDB fits the variant, across roughly 4,000 reachable TP53 variant profiles
- Live measured binding lookup: the Measured binding card queries ChEMBL for the exact variant plus same-residue context, normalises units to nM, and links to the original assay record
- Interactive residue focus: click any sequence position to zoom the 3D viewer with legible atom, residue, and element labels at any scale
- Two-stage screening of curated rescuers plus 3,311 approved drugs (ChEMBL max_phase 4): Stage 1 cavity-fit prefilter, Stage 2 in-browser physics docking, ligand-efficiency ranking, one compound per run with 1/5/10/25/50 add-more increments
- De-novo lane: pocket-sized structures proposed from a fragment/mask seed, filtered to the cavity heavy-atom window, then docked and ranked on the same local physics terms as the curated library; unsynthesised computational proposals, labelled as such
- Chemistry-accurate 3D rendering of every pose: CPK element colours and van der Waals radii, single/double/triple/aromatic bond geometry, wedge and dash stereo cues, element letters, and a ligand-versus-protein scale control, with one compound in the scene by default
- Per-pose ligand-residue interaction map, structure-validation gate before render, and Pose Checks-style geometric sanity checks
- Chemical-space workbench: ECFP4 Tanimoto similarity search and descriptor scaffold-hop over the fingerprint index built in a background worker and cached in IndexedDB
- Exact Chem Toolkit descriptors with PAINS/Brenk structural alerts, a documented 0-100 developability triage index, high-confidence filtering, a 2D PCA map, browser-local notes, and CSV export of the filtered or selected compound set
- Side-by-side comparison of the docked shortlist with Chem Toolkit descriptors in one dialog
- Rescue-vs-inhibit reading with calibrated probability, ensemble spread, and confidence band
- Curated motif attribution and per-residue ΔΔG heuristic for the p53 hotspot variants
- Deep-mutational-scanning overlay on the same residue view as the prediction
- Prime editing sensor evidence: 1,227 variants with in-cell MAGeCK LFC across timepoints and Nutlin-3 selection
- Live disease and homology context with a secondary bioactivity source so upstream outages do not stall the session
- Provenance toggle across every table and plot: measured-only mode withholds modelled and score-derived rows, and combined mode keeps them as a separate, never-merged series
- Dataset quality page: per-file row counts, molecule and structure-identifier coverage, affinity-unit normalisation and pass/fail validation checks for every shipped dataset
- Docking visualisation panel: pose inputs, residue-numbering offsets, ligand provenance, and an overlay measuring centroid offset and per-atom distance against every crystal ligand in the receptor file
- Structured JSON export of the full visible screen: cards, tables, links, scores, and citations, with uncomputed fields left null
Look up a variant. Get catalogue-backed records.
Variant lookups run as deterministic search over the curated mutation catalogue, so every answer maps back to a catalogue row with its citation - no generated prose, no runtime model.
>Curated rescue compounds for R248Q
>TP53 hotspots ranked by dockable-pocket score
>R175H versus R273H mechanism records
Frequently asked, honestly answered.
What it does, how confident it is, and where it stops. Ten chapters, plain answers. Pick one to begin.
Open a variant and work through it end-to-end.
Sign in with your credentials. Accounts are provisioned by the HyperLab53 team.