Drybench Assay.
What the assay analyser does not do
Where a 4PL fit here stops: four concentrations at least, no extrapolation, the interval assumptions, relative against absolute C50 and the 400-curve cap.
What a four-parameter logistic fit can and cannot tell you, and where this app stops. Each limit below is also in the app's Limits panel and in every export's provenance block.
For research and education only, and not medical advice: a C50 estimated here describes this assay under these conditions and is not a binding constant.
A C50 depends on the assay's conditions: the cell line or enzyme, the incubation time, the readout and the normalisation. It is a descriptive number for this experiment, not a binding affinity, and comparing it across assays needs the conditions beside it.
The reported C50 is always the four-parameter logistic fit, with any parameter fixable; no five-parameter dose-response fits (standard curves, ELISA among them, are fitted in the Standard curve analysis). A bell-shaped or biphasic curve can be fitted beside it for one curve, for information: it describes the curve's shape over the tested concentrations, does not show two binding sites, two targets or two mechanisms, does not say why the response turns, and reports no interval.
Bell-shaped and biphasic curves are fitted only when you choose them, beside the four-parameter logistic; the five-parameter logistic is offered for standard curves only, never for dose response, because its midpoint is not a C50; GR metrics have their own analysis. The model used is stated in the method block and the caption.
The app refuses to fit fewer than four distinct concentrations, treats a concentration of zero as a control, and never extrapolates: a C50 the tested concentrations do not bracket is withheld with the reason.
With three concentrations the four parameters are not identified. A zero concentration has no logarithm, so vehicle wells are controls, not points. A curve whose C50 or plateau the tested concentrations do not bracket is reported as not determinable in that range.
Confidence intervals are profile extra-sum-of-squares by default and Wald as the named alternative, both on log10 C50 and both assuming independent errors of constant variance; relative fits assume SD proportional to response and variance proportional to its square, and have no intervals; a bound not reached within ten log units is not bounded within the search.
Both unweighted intervals assume independent Gaussian errors with constant variance and an identifiable fit. Relative IRLS assumes SD proportional to positive fitted response, hence variance proportional to its square (constant CV); it reports no intervals until validated. Repeated readings of one well are not independent, and the 4PL does not model random plate effects or control-mean uncertainty. The unweighted profile uses an approximate extra-sum-of-squares F threshold; Wald is a local linear approximation on the logarithm.
The relative C50 is the curve's own midpoint; the absolute C50 is where the curve crosses 50 percent of the controls, needs responses on a percent-of-control scale and two concentrations predicted on each side of 50, and is not determinable otherwise.
The relative C50 is a parameter of the fitted curve. The absolute C50 needs responses as percent of control and a curve that crosses 50 percent within the tested range, with two concentrations predicted on each side (Sebaugh 2011).
Normalisation uses each plate's own control means, and Z′ uses sample standard deviations: Z′ below 0.5 is an observation against Zhang's convention, not a pass mark, and R² is shown but is not a goodness-of-fit gate.
Z′ measures the separation of the controls on one plate with sample standard deviations; the 0.5 convention is from Zhang, Chung and Oldenburg 1999 and is reported as an observation. R² is shown because readers expect it, and is not used to accept or reject a fit.
Hill slope comments are observations, not diagnoses.
A Hill slope far from one can mean tight binding, aggregation, a mixed population or a normalisation artefact; the app reports the observation and no cause.
GR metrics correct for the division rate within the assay and assume exponential growth and a readout proportional to cell number; a drug that changes signal per cell biases them, GR below 0 means net cell loss rather than a mechanism, GR_AOC compares only across the same concentration range, and they are not checked against the GRmetrics package.
Ligand efficiency divides a potency by the heavy-atom count; for a cell-viability or phenotypic C50, which is not a binding measurement, the ratio has no meaning, so it is not computed.
Kinetic read blocks are checked on synthetic files only, not on an instrument's own export. A dose-response fit takes one read per well, the last or the one you choose, named in the method block and paragraph; growth, progress-curve and melt analyses take every read of the wells you name. Reads are never interpolated or filled in; a missing or reader-marked read is left out with the reason; a temperature column beside time is kept as metadata, not corrected for; pre-corrected or dual-wavelength values are not recombined.
Four hundred curves fit in a worker in a few tens of seconds; a larger import is split before fitting, so the progress and the cancel stay meaningful.
No ligand efficiency is computed for cell-viability or phenotypic C50s.
Plate-reader vendors publish no licence for their export formats, so the profiles were built on files the suite generated, recognised by the plate block's structure, and the header is discarded because it often carries the operator's name, the user id, the instrument serial and the file path.
A batch takes at most 400 curves; a larger import is split before fitting.
Your plate data are unpublished intellectual property, so they never travel in a link: a shared link carries the view and the fit settings only, and the assay lives in this browser and in the session file you export. No account is needed to analyse and export; saving needs one.
Plate layouts cover the plain grid and other recognised layouts at 96 and 384 wells; header metadata, including any user name, instrument serial or file path, is read past and discarded, and any other layout is read as a table with units in its headers.
ChEMBL's bioactivity data are share-alike, so the app links to ChEMBL's page for a target and shows no ChEMBL row.
Your plate data are not placed in links: an assay is kept in this browser as an object and in the session file you export, and a shared link carries the view and fit settings only.
The pages are served by the site each time; nothing is installed and no offline use is claimed.
Literature potencies are links to ChEMBL, never rows in the app or its exports.
No offline use: the pages are served by the site, and nothing is installed.
Version 2 intervals are unavailable for relative IRLS and numerically zero-residual fits. Unweighted intervals condition on normalised observations and omit uncertainty shared through estimated control means. Permutation response evidence assumes exchangeability within plate; technical wells are not biological replicates. Relative IRLS requires positive observed responses and positive fitted predictions throughout; nonpositive responses are refused rather than offset or silently excluded. This is an app domain restriction, not a requirement of all constant-CV estimators.
A compound known only by its label stays in the assay's data; its C50 is written on the target only when the assay names one, and a structure (a SMILES column, or a compound sent from the molecule app) is what lets a value sit on the compound itself.
Reopening an assay under a newer engine version recomputes only on request and shows the differences against the stored results; the stored results stay until you refit.
Reads a plate export or table up to 4.1 MB and opens a project file up to 50.0 MB; a larger file is refused, not cut.
The figure is edited through its settings: text, reference lines, shaded ranges and arrows are anchored to data values, never placed or drawn by hand, at most 50 on a figure; a saved figure style is kept in this browser on this device.
A negative result is a statement about the test, not the compound: no monotonic trend at p < 0.05 within plates; offered only where the design could have reached p < 0.05; no power calculation.