LD50 / LC50 Probit Analyzer
Estimate LD50 or LC50 from grouped mortality data using binomial maximum-likelihood probit analysis. Analyze dose–response relationships, confidence intervals, probit slope, intercept, and model goodness of fit.
LD50 / LC50 Probit Analysis for Research
LD50 and LC50 are commonly used endpoints in toxicology, pesticide research, entomology, pharmacology, environmental toxicology, aquatic toxicity, and mortality bioassays. They describe the exposure level associated with approximately 50% lethality under the specified experimental conditions.
The ResearchUtility LD50/LC50 Probit Analyzer provides a browser-based workflow for analyzing grouped mortality data. Instead of simply selecting the tested dose that happens to produce 50% mortality, probit analysis models the relationship between mortality probability and log-transformed exposure.
LD50 / LC50 Probit Analyzer
Upload your grouped mortality dataset, select the required variables, and perform probit-based LD50 or LC50 analysis.
LD50 / LC50 Probit Analyzer
Estimate LD50 or LC50 from grouped mortality data using binomial maximum-likelihood probit analysis.
What Is LD50?
LD50 means median lethal dose. It is the statistically estimated dose associated with 50% mortality in a defined experimental population under specified experimental conditions.
Depending on the study, the dose may be expressed in units such as mg/kg body weight, µg/animal, mg/organism, or another experimentally appropriate unit.
An LD50 should therefore not be interpreted as the dose at which one particular organism dies. It is a population-level dose–response endpoint estimated from mortality observations.
What Is LC50?
LC50 means median lethal concentration. It represents the concentration associated with 50% mortality under specified exposure conditions.
LC50 is frequently used in aquatic toxicity, environmental toxicology, pesticide studies, chemical toxicity, wastewater testing, and other experiments where exposure is represented as a concentration rather than an administered dose.
What Is Probit Analysis?
Probit analysis is a statistical approach for analyzing quantal responses such as dead/alive, affected/unaffected, or responder/non-responder outcomes.
In mortality studies, the method relates the probability of death to dose or concentration. The exposure variable is commonly represented on a logarithmic scale, while mortality proportions are modeled through the probit link.
Probit(response) = Intercept + Slope × log₁₀(Dose or Concentration)
The LD50 or LC50 corresponds to the exposure level at which the fitted model reaches a 50% mortality probability.
What Data Does the LD50/LC50 Analyzer Require?
The analyzer is designed for grouped mortality data supplied through a CSV file. The dataset should contain the exposure level together with the number of organisms exposed and the number that died.
| Dose / Concentration | Exposed | Dead | Mortality % |
|---|---|---|---|
| 1 | 20 | 0 | 0% |
| 2 | 20 | 1 | 5% |
| 4 | 20 | 4 | 20% |
| 8 | 20 | 10 | 50% |
| 16 | 20 | 17 | 85% |
| 32 | 20 | 20 | 100% |
The exact column names in your CSV may differ. During analysis, select the columns corresponding to dose/concentration, number exposed, and number dead.
What Does the Probit Analyzer Calculate?
Understanding the 95% Confidence Interval
The confidence interval provides information about uncertainty surrounding the estimated LD50 or LC50. A point estimate alone does not communicate how precisely the endpoint has been estimated.
LD50 = 7.39 mg/kg
95% CI = 5.88–9.30 mg/kg
The LD50 should be reported together with its units and, where appropriate, the corresponding confidence interval.
A very wide confidence interval can indicate substantial uncertainty in the estimated endpoint. Confidence intervals should therefore be considered when comparing toxicity estimates between experiments.
Understanding the Probit Slope
The probit slope describes how rapidly the modeled probability of mortality changes with increasing log10 dose or concentration.
A steeper slope represents a more rapid transition from lower to higher predicted mortality across the exposure range. The slope should be interpreted together with the fitted curve, dose spacing, sample size, and experimental design.
Probit = Intercept + Slope × log₁₀(Exposure)
The intercept is a mathematical parameter of the fitted model and generally should not be interpreted as an independent biological endpoint.
Understanding Pearson χ² Goodness of Fit
The analyzer reports a Pearson chi-square statistic and corresponding goodness-of-fit p-value to help assess whether the fitted probit model adequately describes the observed grouped mortality data.
Suppose Pearson χ² = 0.83, df = 4 and p = 0.934.
At the 0.05 significance level, this result does not provide statistically significant evidence of lack of fit.
A non-significant goodness-of-fit result does not prove that the model is correct or that the model is biologically perfect. It means that the specified goodness-of-fit test did not detect statistically significant evidence of lack of fit at the chosen significance level.
Understanding the Probit Mortality Curve
The dose–response curve helps visualize the relationship between exposure and mortality. The analyzer displays mortality against log10 dose or concentration and identifies the 50% response level used to estimate LD50 or LC50.
Inspecting the graph is important because a numerical estimate should not be interpreted independently of the observed data. The fitted curve should be broadly compatible with the mortality pattern across the tested exposure range.
Toxicity experiments often use concentrations or doses spanning several orders of magnitude. A logarithmic scale makes these exposure levels easier to visualize and is commonly used in dose–response analysis.
Why Partial Mortality Groups Matter
Partial mortality groups are exposure groups where mortality is greater than 0% but less than 100%.
Intermediate mortality groups provide important information about the transition from low to high mortality and are particularly important when estimating the midpoint and confidence limits of a dose–response relationship.
If the tested exposure levels contain only 0% and 100% mortality, the location of the 50% response may be poorly constrained. For a more informative analysis, the experimental dose range should generally include meaningful intermediate mortality responses.
How to Use the LD50 / LC50 Probit Analyzer
Prepare your CSV dataset
Prepare grouped mortality data containing dose or concentration, number exposed, and number dead.
Upload the CSV file
Upload your dataset using the analyzer’s CSV input.
Select the analysis type
Choose whether the exposure should be analyzed as a dose for LD50 or a concentration for LC50.
Select the required variables
Identify the dose/concentration, exposed count, and dead count columns.
Run Probit Analysis
Start the analysis to fit the binomial maximum- likelihood probit model.
Review the results
Examine LD50/LC50, confidence interval, slope, intercept, goodness of fit, and the fitted curve.
Download and document your results
Download the curve or results and retain the experimental data and analysis settings for reproducibility.
Example of LD50 Probit Analysis
Consider a hypothetical mortality experiment with the following dose-response pattern:
| Dose | Exposed | Dead | Observed Mortality |
|---|---|---|---|
| 1 mg/kg | 20 | 0 | 0% |
| 2 mg/kg | 20 | 1 | 5% |
| 4 mg/kg | 20 | 4 | 20% |
| 8 mg/kg | 20 | 10 | 50% |
| 16 mg/kg | 20 | 17 | 85% |
| 32 mg/kg | 20 | 20 | 100% |
The analyzer does not simply select 8 mg/kg because this tested dose produced 50% observed mortality. Instead, the fitted statistical model uses the information from the complete dose-response dataset to estimate the exposure corresponding to 50% mortality.
If the fitted model estimated an LD50 of 7.39 mg/kg with a 95% confidence interval of 5.88–9.30 mg/kg, the estimated LD50 would be reported as 7.39 mg/kg under the specified experimental conditions.
LD50 vs LC50: What Is the Difference?
| Feature | LD50 | LC50 |
|---|---|---|
| Meaning | Median lethal dose | Median lethal concentration |
| Exposure variable | Dose | Concentration |
| Response | Mortality | Mortality |
| Common applications | Toxicology, pharmacology, pesticide research | Aquatic and environmental toxicity |
LD50 Is Not the Same as EC50
LD50 and LC50 specifically concern lethality. EC50 refers to the exposure producing 50% of a specified effect and does not necessarily mean that organisms die.
For example, EC50 may be used for growth inhibition, enzyme response, behavioral response, pharmacological response, or another defined biological endpoint.
If your experimental response is mortality, LD50 or LC50 may be appropriate depending on whether the exposure is expressed as dose or concentration. If the response is non-lethal, another dose-response endpoint may be more appropriate.
Important Data-Quality Checks Before Probit Analysis
- Confirm that all dose or concentration values are positive.
- Verify that the number dead does not exceed the number exposed.
- Check that dose or concentration units are consistent.
- Check for transcription errors in mortality counts.
- Confirm that each row represents the intended experimental exposure group.
- Check whether mortality responses provide information around the 50% region.
- Review the fitted curve rather than relying only on the numerical LD50/LC50 estimate.
- Record the experimental conditions, exposure duration, organism, substance, and units when reporting the result.
Common LD50/LC50 Probit Analysis Mistakes
1. Treating LD50 as an observed dose
A model-derived LD50 does not have to equal one of the doses actually tested. It can fall between experimental dose levels.
2. Reporting LD50 without units
A result such as “LD50 = 7.39” is incomplete. The dose unit should be reported with the numerical estimate.
3. Ignoring the confidence interval
The point estimate alone does not describe uncertainty. Report the confidence interval when it is estimable and appropriate for the analysis.
4. Assuming p > 0.05 proves the model
A non-significant goodness-of-fit result does not prove that the fitted model is correct. It indicates that the specified test did not detect statistically significant evidence of lack of fit.
5. Ignoring the dose range
An LD50 estimated far outside the tested exposure range may involve substantial extrapolation and should be interpreted cautiously.
6. Using LD50 for a non-lethal endpoint
If the measured response is inhibition, growth, activity, development, or another non-lethal endpoint, an appropriate effect endpoint should be considered instead.
Research Interpretation of LD50/LC50 Results
An LD50 or LC50 should be interpreted within the biological and experimental context in which it was obtained. The numerical endpoint depends on factors such as organism, strain, developmental stage, exposure route, exposure duration, formulation, environmental conditions, dose spacing, and experimental design.
Two experiments can produce different LD50 values even when the same substance is studied if their experimental conditions differ.
• Test organism and biological stage
• Test substance and formulation
• Exposure route
• Exposure duration
• Dose/concentration units
• Number of organisms per group
• Mortality observations
• Statistical model
• LD50/LC50 estimate
• 95% confidence interval
• Probit slope
• Goodness-of-fit statistic
How to Report LD50 Results in a Research Paper or Thesis
A concise statistical reporting format may include the model, estimated lethal endpoint, confidence interval, slope, and goodness-of-fit information.
LD50 = 7.39 mg/kg (95% CI: 5.88–9.30 mg/kg)
A complete report should also describe the organism, test substance, exposure conditions, sample size, and statistical method used to obtain the estimate.
When preparing a thesis, dissertation, manuscript, or research report, retain the original mortality dataset so that the analysis can be independently checked or reproduced.
Limitations of Probit Analysis
Probit analysis is a useful dose-response method, but the quality of its output depends on the experimental data and statistical assumptions.
- The response must be appropriately represented as a quantal or grouped mortality response.
- Dose or concentration values must be suitable for logarithmic transformation.
- Extremely limited mortality variation can make estimation of the midpoint and confidence interval difficult.
- Poorly selected exposure ranges can lead to unstable estimates or substantial extrapolation.
- A numerical result should not be interpreted independently of biological and experimental conditions.
- Alternative dose-response models may be appropriate for some datasets.
This analyzer is a research calculation and analysis aid. For regulatory toxicology, publication, or high-stakes scientific decisions, verify the analysis against the applicable study protocol, statistical methodology, and regulatory guidance.
Privacy-Friendly Browser-Based Analysis
The current ResearchUtility analyzer processes the uploaded CSV locally in the browser rather than sending the dataset to ResearchUtility’s server.
This can be useful for researchers working with unpublished laboratory datasets, thesis data, preliminary experiments, and mortality bioassay results.
Researchers should nevertheless follow their institution’s requirements for handling confidential, proprietary, or otherwise sensitive research data.
Best Practices for LD50/LC50 Analysis
Frequently Asked Questions
What is an LD50?
LD50 is the estimated dose associated with 50% mortality in a defined experimental population under specified conditions.
What is an LC50?
LC50 is the estimated concentration associated with 50% mortality under specified experimental conditions.
What is probit analysis?
Probit analysis is a statistical method used to model quantal responses such as mortality as a function of dose or concentration.
Can I calculate LD50 from a CSV file?
Yes. The ResearchUtility analyzer accepts grouped mortality data in CSV format containing exposure level, number exposed, and number dead.
Does the LD50 have to equal one of my tested doses?
No. A model-derived LD50 can fall between tested dose levels because the fitted model estimates the exposure corresponding to a 50% response.
Why are partial mortality groups important?
Mortality groups between 0% and 100% provide information about the transition from low to high mortality and can improve estimation of the dose-response relationship.
What does the 95% confidence interval mean?
The confidence interval communicates uncertainty around the estimated LD50 or LC50 under the statistical method used.
What does a high goodness-of-fit p-value mean?
It means that the specified goodness-of-fit test did not detect statistically significant evidence of lack of fit at the selected significance level. It does not prove that the model is correct.
Can this analyzer be used for aquatic toxicity?
Yes. LC50 analysis is commonly used for aquatic and environmental toxicity studies when concentration-based mortality data are appropriate for the analysis.
Can I use this analyzer for pesticide bioassays?
It can be useful for appropriate pesticide mortality bioassay datasets. Researchers should follow the applicable experimental and regulatory methodology for their specific study.
Can I use this tool to calculate EC50?
Not automatically. EC50 concerns a specified biological effect and is not synonymous with lethal mortality. Use an analysis appropriate to the response endpoint.
Why must dose or concentration values be positive?
The analyzer uses log10 dose or concentration in the probit relationship. Logarithms of zero or negative exposure values are undefined.
Related Research Data Analysis Tools
Continue your research workflow with other ResearchUtility data analysis tools for cleaning, transformation, exploration, visualization, and statistical analysis.
Analyze Your LD50 or LC50 Data
Use the ResearchUtility LD50/LC50 Probit Analyzer to analyze grouped mortality data, estimate the median lethal endpoint, review confidence intervals and inspect the fitted dose-response relationship.
Start Probit Analysis