Statistical analysis: Available tests and how to run them

BioRender Graphing provides a comprehensive suite of statistical analysis tools to help you interpret your experimental data with confidence. This guide outlines the available analyses and how to run each type.

Not sure which test to run? When possible, Graphing’s analysis guidance recommends the most appropriate test based on your data - running normality and equal variance checks in the background so you don’t have to look them up or ask an expert. If you choose a different test, BioRender will explain the trade-off so you can run analyses consistently, every time.

Table of contents

• How analysis guidance works
• t-tests
• One-way ANOVAs
• Two-way ANOVAs
• Multiple comparisons tests
• Linear regressions
• Dose-response regression
• Survival analysis
• Logistic regression
• Correlation

How analysis guidance works

BioRender automatically runs two assumption checks for categorical test in the background (t-tests and ANOVAs):

The recommended test is marked with a badge in the analysis panel. You can override it at any time - if you select a test that doesn’t match your data’s assumptions, BioRender will explain why it’s still recommending the original test so you understand the trade-off. Some tests may appear greyed out if your data doesn’t meet the requirements - hover over the tooltip to understand why.

📌 Note: These suggestions are guidelines. Always consider the conventions of your field when making your final decision.

T-tests

Use t-tests to compare differences between two groups. BioRender offers the following options:

• Unpaired t-test
• Unpaired t-test with Welch’s correction (Welch’s t-test)
• Paired t-test
• Mann-Whitney U test (nonparametric)
• Wilcoxon matched-pairs signed ranks test (nonparametric)

Setting up a t-test

1. On the Graph editing page, click + Run new analysis.
2. If your dataset has two groups, a t-test will be suggested. Descriptive statistics (mean, median, standard deviation, etc.) are automatically calculated with every test.
3. Use the analysis panel to select the appropriate test. The recommended test is marked with a badge - hover over it for an explanation of why it was suggested.
4. Fine-tune your analysis in Step 2.
5. Choose your hypothesis testing approach (P. value) in Step 3.
6. Click Run. Results appear in the Analysis panel, and significant comparisons are automatically marked on your graph.

One-Way ANOVAs

Use a one-way ANOVA to compare differences between three or more groups. Available options:

• One-way ANOVA
• Welch’s one-way ANOVA
• Kruskal-Wallis test (nonparametric)

Setting up a one-way ANOVA

1. On the Graph editing page, click + Run new analysis.
2. If your dataset has three or more groups, a one-way ANOVA will be suggested. Descriptive statistics are automatically reported with the analysis result.
3. BioRender will recommend a suitable test, marked with a badge. Hover over it for an explanation.
4. If you would like to change your preferred test, select your test and configure any additional settings.
5. Fine-tune your Analysis in Step 2.
6. Choose groups for your multiple comparisons test in Step 3
7. Click Run Analysis.

💡 Tip: If your ANOVA result is significant, you’ll likely want to run a multiple comparisons test to identify which specific groups differ. See the Multiple Comparisons section below.

Two-Way ANOVAs

Use a two-way ANOVA to analyze the interaction between two independent variables.

Setting up a two-way ANOVA

1. On the Graph editing page, click + Run new analysis.
2. Select two-way ANOVA from the analysis type dropdown.
3. Select the groups you want to compare (minimum of 2 per independent variable) and choose your experimental design in Step 2.
4. Decide whether to run multiple comparisons tests in Step 3. If yes, choose which groups to compare and select your post-hoc test. BioRender suggests an appropriate test based on your data, marked with a badge.
5. Click Run Analysis.

Multiple comparisons tests

After a significant ANOVA result, multiple comparisons tests help identify which specific groups differ. BioRender suggests an appropriate post-hoc test based on your data and comparison type.

Here are all of the multiple comparison tests offered:

To run a multiple comparisons test, click on this option in Step 3 in the analysis panel when setting up your ANOVA. Choose whether to compare all groups against each other or against a control group - BioRender will suggest the appropriate test based on your selection.

Linear Regression

Use linear regression to analyze the relationship between two continuous variables.

Setting up a linear regression

1. On the Graph editing page, click + Run new analysis.
2. Select your regression model (simple or with a grouping variable).
3. Decide whether to constrain the Y-intercept to zero based on your experimental setup.
4. Click Run. The best-fit line is automatically displayed on your graph along with the regression output.

📌 Note: Multiple regression is not currently available, but it is coming soon. If you have a categorical variable in your plot legend, a group-wise univariate regression will run automatically.

Dose-Response Regression

Also known as dose-response curves, Hill curves, Hill equation, or EC50/IC50 analysis. Use this to analyze the effect of varying doses on a measured response.

Setting up a dose-response regression

1. On the Graph editing page, click + Run new analysis.
2. Select the type of experiment: Stimulation or Inhibition, then click Continue.
3. Specify whether you’ve already log-transformed your X values using base 10.
4. Decide whether to constrain your Hill slope to 1, depending on your experimental setup.
5. Click Run. BioRender fits the best-fit curve automatically and displays it on your graph. 

Survival analysis

Also known as duration or event history analysis, this is used to estimate the probability of an event occurring over time.  Analyses of this data include the Kaplan-Meier estimation and Cox proportional hazards regression.

Setting up a survival analysis

1. On the Graph editing page, click + Run new analysis.
2. Select which groups you want to analyze
3. Select which comparison method or methods you want to use
4. Select which multiple comparison correction you want to use
5. Choose a method for handling tied event times
6. Click Run Analysis. BioRender fits the model and plots the curve on your graph.

Logistic Regression

Also known as a logit regression. Use this analysis to predict the outcome of a binary event (e.g. alive/dead, yes/no)

Setting up a logistic regression

1. On the Graph editing page, click + Run new analysis.
2. Click Run Analysis. BioRender fits the model and plots the curve on your graph.

Correlation

Use this analysis to find how two variables are related to one another. Are higher values for one variable associated with higher or lower values of the other variable?

Setting up a correlation analysis

1. On the Graph editing page, click + Run new analysis.
2. Select whether you want to do Pearson or Spearman correlation
3. Click Run Analysis. BioRender calculates the correlation and plots the line on your graph.

Summary

• t-tests – Comparing two groups
• One-way ANOVA – Comparing three or more groups
• Two-way ANOVA – Analyzing interaction between two independent variables
• Multiple comparisons – Identifying which groups differ after an ANOVA
• Linear regression – Assessing the relationship between two continuous variables
• Dose-response regression – Evaluating dose-response relationships
• Survival analysis – Estimating the probability of an event occurring over time
• Logistic regression – Predicting the outcome of a binary event
• Correlation – Measuring how two variables are related to one another