Models

Inputs. Atomic elements and coordinates, electronic-structure settings, or a molecular system with positions, masses, charges, and force parameters.

Calculations. The electronic-structure solver iterates an electron-density solution for a specified geometry. The molecular dynamics engine advances atomic positions under the chosen force model. Electronic structure and molecular dynamics have separate solver configurations.

Outputs. Electronic-structure results, molecular trajectories, and energy and runtime reports, depending on the selected solver.

Availability. Standalone native solver APIs and a live-atom runtime, with compute backend and system size selected per run.

Inputs. A chemical structure written as text in SMILES notation, ligand coordinates and a binding-site representation, or a dosing schedule and physiological parameters.

Calculations. Molecule-building and docking tools explore poses with approximate scoring. Compartment models track absorption, distribution, and elimination. A selected whole-body workflow adjusts clearance using liver and kidney state, models competition between drugs for metabolic enzymes, and passes free drug concentrations to an optional heart-cell model.

Outputs. Molecular descriptors, candidate poses and approximate scores, concentration trajectories, and modeled exposure changes.

Availability. Standalone molecule and docking tools, plus selected exposure-to-physiology coupling. The docking tools use simplified binding-site proxies.

Inputs. Nutrient and energy conditions, pathway stimuli, and cell-model configuration. For example, the growth-signaling model accepts growth factors, amino-acid availability, an energy-stress signal, and rapamycin exposure.

Calculations. Pathway models evolve metabolism and signaling states. The whole-cell simulator, configured around a minimal bacterial cell (Syn3A-like), schedules chemical reaction and diffusion, expression and assembly, flux calculations, replication, and cell geometry. Selected local atomistic domains can be enabled separately.

Outputs. Metabolite pools and pathway activity; whole-cell reports also include active transcription and translation machinery, replicated bases, geometry, and division progress.

Availability. Standalone pathway APIs and an optional whole-cell engine, with local atomistic domains enabled separately.

Inputs. Tissue geometry, vessel positions, nutrient boundaries, and local cell viability or tissue state.

Calculations. A three-dimensional tumor-microenvironment model evolves oxygen, glucose, and lactate with spatial transport and metabolism. It derives a pH estimate from lactate. Separate models cover reaction–diffusion patterns and spatial liver regeneration.

Outputs. Spatial chemical fields, pattern wavelength and amplitude, and tissue-level observables.

Availability. Standalone spatial models, with an opt-in liver-tissue connection to organ physiology.

Inputs. Physiological starting state, ventilation and circulation settings, metabolic demand, disease or intervention parameters, and explicitly enabled submodels.

Calculations. The Heart–Brain–Lung simulator advances breathing, blood gases, circulation, and feedback loops. Slower updates step or couple organ processes, including coronary metabolism, liver glycogen, and kidney feedback. Four-chamber circulation, ventricular cells, and other detailed components have their own activation paths.

Outputs. Time-varying blood gases, oxygen saturation, pressure, ventilation, organ-state measurements, and modeled responses to interventions. Separate cross-scale examples trace selected molecular perturbations into downstream physiological outputs.

Availability. A coupled native core with optional components. Current cardiomyocyte work includes the Nguyen et al. (2017) mechanical-assay reproduction; see its current validation status.

Inputs. Neurons and directed connections, optional morphology, stimulation currents, neurotransmitter conditions, and modeled drug exposure.

Calculations. The neural engine steps membrane and synaptic activity. Its pharmacology model evolves estimated exposure and receptor occupancy, then changes neural conductances. Its pharmacology implementation is separate from the whole-body exposure model.

Outputs. Membrane state, spikes, firing rates, and local-field projections at chosen probe positions. These let a researcher compare circuit response under specified changes.

Availability. Native circuit constructors and simulation APIs, with optional mechanisms and backend selection.

Inputs. Host chemical and physical boundaries, such as hormones, oxygen, temperature, and local mechanical loads. A separate C. elegans embryo model accepts temperature and simulation time.

Calculations. The reproductive host coordinates endocrine, ovarian, uterine, vaginal, and transport models. The worm embryo model advances cell polarity, neighboring-cell signals, replication, fate tendencies, and division.

Outputs. Reproductive tissue and hormone panels, transport reports, or developing cell states and counts.

Availability. Native reproductive-host and separate C. elegans developmental APIs. The reproductive host covers tissue physiology and transport; fertilization and fertility outcomes are outside that host’s scope.

Inputs. Sensory conditions for an organism model, or initial populations, nutrient availability, and environmental settings for an ecological scenario.

Calculations. Organism-level APIs include a sensory-driven C. elegans model. An independent pond model couples algae, daphnia, and fish populations with nutrient-limited growth, predation, mortality, and biogeochemical cycling.

Outputs. Organism state and response, or ecological population trajectories and chemical pools. The pond model can compare nutrient-limited and other specified scenarios.

Availability. Native organism models and an independent ecosystem library. The pond integrator runs separately from the experimental Terrarium system.