Nephron Lab: an interactive renal physiology simulator
Nephron Lab is a free, browser-based teaching model of how the kidney filters blood. Change the physiology and watch a single nephron respond: glomerular filtration rate, estimated GFR, creatinine, urea, urine output and sodium handling all update over simulated days.
Illustrative model for teaching. Not patient-specific, and not for clinical decisions.
What you can change
- Blood flow: mean arterial pressure, afferent and efferent arteriolar tone, Bowman's space pressure, venous congestion, nephron number, autoregulation
- Intake: fluid balance, sodium intake, protein intake
- Hormones: angiotensin II, aldosterone, ADH (vasopressin), atrial natriuretic peptide
- Drugs: ACE inhibitors and ARBs, NSAIDs, SGLT2 inhibitors, loop diuretics, thiazides, mineralocorticoid receptor antagonists
Clinical scenarios
- Prerenal acute kidney injury from volume depletion, with a fractional sodium excretion below 1% and a raised urea-to-creatinine ratio
- NSAIDs in a dehydrated patient, removing the prostaglandin support that keeps the afferent arteriole open
- ACE inhibitor in bilateral renal artery stenosis, where filtration depends on efferent constriction
- The triple whammy: ACE inhibitor or ARB, plus a diuretic, plus an NSAID
- Starting an SGLT2 inhibitor and the expected early dip in eGFR from tubuloglomerular feedback
- Nephron loss with compensatory hyperfiltration
- Ureteric obstruction raising Bowman's space pressure
- Haemorrhage with hypotension
How it works
Filtration is modelled along the glomerular capillary using the classic single-nephron Starling-force framework, with oncotic pressure from the Landis–Pappenheimer relation. Creatinine and urea are tracked over time, so estimated GFR visibly lags true GFR after an acute change — the reason eGFR misleads in acute kidney injury. Estimated GFR uses the race-free CKD-EPI 2021 creatinine equation.
Every parameter carries its unit, range and source, and the model documentation states which values are verified against published figures, which are model choices, and where the model knowingly departs from the literature.
Source code and model documentation on GitHub
Loading the interactive simulator… it needs JavaScript enabled.