Integumentary
Skin and associated structures; protection and temperature regulation.
Scientific reference ↗Loading the reference index…
Select a record.
ChEBI: checked three-star entries, with formula, charge, average mass, and monoisotopic mass where available. Reactome: unique human event identifiers from the event hierarchy. These snapshots are imported, not refreshed live.
ChEBI / CC BY 4.0 ↗ · Reactome data license ↗Navigate biological scale: systems → organs → tissues → cells → organelles → molecules. These reference summaries organize the research library; they do not constitute a complete body model.
Skin and associated structures; protection and temperature regulation.
Scientific reference ↗Bones and joints; support, protection, and mineral storage.
Scientific reference ↗Skeletal muscles; movement and heat production.
Scientific reference ↗Brain, spinal cord, and nerves; rapid coordination.
Scientific reference ↗Hormone-producing glands; chemical regulation.
Scientific reference ↗Heart and vessels; circulation and transport.
Scientific reference ↗Lymphatic structures; fluid return and immune defense.
Scientific reference ↗Airways and lungs; gas exchange.
Scientific reference ↗Digestive tract and accessory organs; nutrient processing.
Scientific reference ↗Kidneys and urinary tract; waste and fluid balance.
Scientific reference ↗Reproductive organs; gametes and reproductive functions.
Scientific reference ↗Fluid environment surrounding intracellular structures.
Scientific reference ↗Organelles involved in cellular energy conversion.
Scientific reference ↗Sites of protein synthesis.
Scientific reference ↗Membrane network involved in protein and lipid processing.
Scientific reference ↗Processes and sorts molecules for transport.
Scientific reference ↗Compartments involved in intracellular breakdown.
Scientific reference ↗Organelles supporting selected oxidative reactions.
Scientific reference ↗Structural network involved in shape and intracellular movement.
Scientific reference ↗Explore tissue expression and subcellular localization through the Human Protein Atlas. Protein-level measurements have not been imported into this snapshot.
Human Protein Atlas ↗Measurements need a biological context, units, method, and source. There is no single universal size or concentration for every human cell.
Water inside cell membranes; potassium and phosphate are important constituents.
Scientific reference ↗Extracellular fluid between tissue cells.
Scientific reference ↗Fluid compartment within blood vessels; contains proteins and dissolved solutes.
Scientific reference ↗Examples include cerebrospinal, synovial, and lymphatic fluids.
Scientific reference ↗Textbook illustrations place adult body water around 50–60% of mass, with variation by age and composition.
Scientific reference ↗Osmosis and membrane transport govern movement between compartments.
Scientific reference ↗A textbook example places a typical animal cell around 10–20 µm in diameter. This is an approximate illustration, not a range that describes every human cell.
NCBI Bookshelf / cell microscopy ↗Length (µm), area (µm²), volume (µm³), and shape.
Concentration (mol/L), mass (g), amount (mol), pH, and compartment.
Temperature, pressure, flow, chronological age, and sampling time.
These are proposed model fields. No individual measurements or clinical reference intervals are supplied.
Explore time-dependent biology and the networks that process energy and materials. Chronological age alone does not define a person's cellular state.
NIH aging research investigates molecular, cellular, and physiological changes over time. Senescence and epigenetic changes are active research topics, with effects that depend on biological context.
NIH / Division of Aging Biology ↗Record age at sampling, tissue, developmental context, model assumptions, and repeated measurements where available.
Reactome organizes human biological events into pathways. Search the imported hierarchy for glucose metabolism, lipid metabolism, amino acids, and energy-related processes.
Reactome / glucose metabolism ↗Cells contain water, inorganic ions, carbohydrates, lipids, proteins, and nucleic acids. Their interactions support structure, signaling, transport, and metabolism. Chemical identity is separate from a molecule's concentration or role in a particular tissue.
NCBI Bookshelf / molecular composition ↗Explore the vision: model a baseline biological state, investigate possible internal futures, trace cellular changes, and compare prevention hypotheses.
Each possible future is conditional on model assumptions. The aim is to identify questions that can be investigated independently, rather than predict a person's exact disease date.
Study selected genetics, epigenetics, molecular signaling, metabolism, immune activity, and cellular processes within a defined model.
Explore ensembles over hours, days, months, and years. Millions or billions of trajectories are aspirations, not existing capabilities.
Compare pathways and preventive intervention hypotheses, including vaccine-related ideas, inside the sandbox.
Prepare a local research plan, inspect an illustrative cellular history, or read the project foundations.
Planning tickets stay in this browser. They are not submitted to a server or used to execute a simulation.
Create a research planning ticket using synthetic inputs. This organizes a proposed experiment; it does not run a biological simulation.
A conceptual lineage illustrates how a future simulator could retain history and trace a flagged state backward. These events and years are invented for explanation.
Define the starting state and model assumptions.
A hypothetical change alters the conditions around a cell population.
An illustrative lineage retains a selected state change.
Track the subsequent event and its preceding state.
Reconstruct the chain of simulated events for investigation.
Compare the lineage with alternative runs and baseline assumptions.
Alter a defined pathway parameter and compare modeled outcomes.
Identify independent data and experiments that test the hypothesis.
The proposed infrastructure combines simulation workers, AI-assisted analysis, databases, and a centralized ticket queue.
Version inputs, assumptions, parameters, and provenance.
Specify horizon, ensemble size, seed, resolution, and resource limits.
Use reproducible workers and lineage event storage.
Summarize modeled trajectories, uncertainty, sensitivity, and failure counts.
Estimate arithmetic throughput under hypothetical sequential operation, with no downtime or overhead.
AI may help identify patterns and select additional simulations. Those proposed functions do not replace biological validation.
The concept, scope, and proposed architecture are documented in your GitHub repository.
Updated October 9, 2026 · Data readiness checked · Clinical prediction not evaluated
A public, de-identified GDC case records height 173 cm and weight 63.3 kg under “Initial Diagnosis.” Its age-at-index field is 61, but that field does not establish the age at the measurement. The full public TCGA-COAD snapshot contains 461 cases; 232 have paired height and weight, all labeled “Initial Diagnosis,” with none labeled “Prior to Diagnosis.” This describes the retrieved records; it does not prove that earlier measurements were never taken.
GDC case and source measurements ↗
The public NLST tables contain 53,452 participant rows and 2,150 cancer records. We found 1,163 first-cancer records linked to an earlier negative baseline screen. One example had screening at day 0 and a lung-cancer diagnosis at day 795. Negative screening does not reveal the condition of every cell, and diagnosis time is not the time of the first mutation.
The public subset omits height and weight. The colon-cancer measurements belong to a different person and are kept separate; there is no height or weight match between these cases.
NLST data and access requirements ↗ · Screening field definitions ↗
A model appropriate to the tissue, pre-diagnosis inputs, measured cell states and mutation rates, a supported calendar-time mapping, and independent validation data. The current colon-crypt model does not apply to lung tissue. No cancer prediction has been run, and no parameters have been adjusted to force the known outcome.
Full NLST and PLCO clinical datasets require approved research access and completed data agreements.
Local research update · October 9, 2026
Private body files preserve supplied measurements, alternatives and sources. The engine counts combined cases exactly and evaluates bounded batches. A colonic-crypt model simulates division and inherited hit states; a separate aggregate crypt-burden model uses supplied fixation and fission rates across five chronological years. These models represent different scales and are not silently connected. A supplied cell-volume model now follows fluid and solid subvolumes with constant targets and rates in minutes. It does not establish five-year cell-size evolution.
Mathematical reference checks, repeatable seeds, implementation hashes and synthetic workflow tests support reproducibility. A model-limit review is distinguished from a simulated two-hit event. Account save/reopen checks preserve model inputs and block access from another account. These checks do not establish clinical predictive accuracy.
The public pre-AML study supplies blood measurements recorded before diagnosis. Its adjusted discovery fit reproduces all 21 published coefficients, and its combined fit reproduces all 25 at printed precision using the pinned author code. A local blood-profile comparison calculates relative hazard. All 796 participant-excluded fits completed. The numerical audit exposed slight increases introduced by the author’s smoothing step; those curves are not ready for probability use. A separate unsmoothed research extension passed numerical checks for all 796 participants, including first-visit-only curves. This does not establish probability calibration. A separate shared-assay research model, trained on 505 discovery participants, was checked on 291 study-validation participants with 29 AML events and no shared source identifiers. Its ranking concordance was 0.773 (standard error 0.062). This is a different model from the full published fit and the dashboard’s rounded-coefficient comparison; it does not validate personal probabilities. Independent prospective deployment validation remains unperformed.
Time-updated author curves can use later visits. A separate baseline-only research check holds first-visit measurements fixed. Neither calculation identifies a particular cell, predicts its first mutation date, or establishes a personal cancer diagnosis.
Published pre-AML study · Author data and code
The published longitudinal blood-clone data have been encoded into 2,789 observations across 303 participants and 685 clones. A registered split keeps 245 participants for training and 58 held out. The trajectory and read-count calculations passed independent numerical checks, and the shared model sampled in a synthetic runtime test.
The first noise-calibration fit finished but failed its convergence gate: maximum Rhat 1.027 and minimum effective sample count 211. Its chains and failed diagnostics are preserved. The longer second attempt passed every preset convergence check (maximum R̂ 1.0061; minimum effective samples 803). Population fitting is now running; held-out evaluation has not been performed. The source contains 12 fractional adjusted X-chromosome counts. The research reproduction preserves the author’s gamma-function approximation and labels it explicitly; strict integer-count mode rejects these values. This approximation is not a validated discrete-count likelihood.
Of 283 later held-out observations, 115 fall within five years of the last starting measurement; only 32 of those involve sites seen in training. Results for unseen sites and longer intervals must be reported separately. This work models variant allele fractions and sequencing read counts, not a particular cell or a personal cancer diagnosis.
Published longitudinal study · Author model code · Public source data
The author-linked HSC simulator release v4.3.13 passed 35 native tests in an isolated, pinned Rust runtime. A versioned research runner validates supplied inputs and records seeds, source and binary hashes, population counts and actual terminal times within Collect → Run properties → Age. A synthetic five-year case preserved 100 cells and ended at 5.0011697 years. The source can cross its requested time limit, and its directory labels round timestamps; these are not exact five-year endpoints.
Independent numerical checks passed for 2,999 sampled fitness rates and 100,000 isolated clone-arrival decisions. These verify implementation rules, not biological forecasts. A deliberately fast synthetic case stopped at 0.50659096 years because of the iteration limit; the runner preserved the failure instead of reporting five-year completion.
The original source’s bounded clone-slot search failed in 75 of 200 trials despite one slot remaining free. A separately versioned repair selects uniformly from actual free slots. All 37 native tests passed, including near-capacity and uniform-selection checks. The repair changes random-number consumption and remains a separate research version; it has not replaced the author release or the dashboard engine.
A separate native diagnostic found that repeated background-mutation updates at the same timestamp add mutations again despite no elapsed time: 10,000 synthetic cells gained a further mean 10.0207 mutations on the second call. Background-burden outputs from both source versions are preserved but quarantined from biological validation. The clone-slot repair does not fix this accounting issue. A separately versioned background-clock repair (hsc/4.3.13-hbs.2) passed 41 native tests and a fixed 100-run neutral mean-burden check; the original source versions remain quarantined.
A separate exact-time Age driver saves eleven half-year observations through year five and skips division events beyond the horizon. Its fixed 100-run neutral ensemble passed the mathematical mean-burden check: mean residual −0.02334 mutations, within the preset tolerance of 0.89610. Boundary and exhausted-budget tests, plus executable failure-preservation checks, passed. These are numerical research checks; selected competition, independent biological validation and dashboard integration remain unfinished.
The Mitchell donor metadata contains 3,592 rows, including 15 exact duplicate sample IDs, leaving 3,577 distinct IDs versus 3,579 colonies reported in the paper. Two donors have multiple timepoint labels but no dated intervals in this metadata. Exclusions and follow-up timing remain unresolved. These data were also used by the newer competition study, so their reuse is not automatically independent validation.
Published competition study · Pinned author simulator release · Donor study · Public donor data
The author’s normal-colon APC tables repeat crypt denominators across variant rows. One record identifier has conflicting ages and three have conflicting denominators. The measurement adapter also rejects 22 strip rows with unusable denominators. Source records and failed checks are preserved; conflicted rows are quarantined. Read counts require an explicit measurement model before comparison to crypt burden.
Normal-colon source data and code · Published model study
Independent biological validation, complete model-specific pre-diagnosis assays, tissue-specific links and supported size-evolution equations remain missing. Two persistent local workers matched serial results on a small synthetic benchmark; this does not demonstrate fast execution of ten billion cases, and the dashboard still evaluates serial batches.
Expand the project's capabilities only as implementation and independent evidence support them.
Vision, boundaries, architecture, research website, and planning dashboard.
Reproducible ensemble runner, lineage events, and reports using synthetic data.
Backend job submission, cancellation, execution, and report retrieval.
Select a subsystem with domain experts and document scientific sources and parameters.
Independent evidence, sensitivity analysis, uncertainty, and measured computational requirements.
Study counterfactual hypotheses with domain researchers and appropriate experimental validation.