#063 — Shapiro-Wilk Normality Test
Statistical Definition
Tests the null hypothesis that a continuous sample is drawn from a normally distributed population using order-statistic correlation.
Bioinformatics Use Case
Pre-flight quality control on qPCR Ct values or log-transformed proteomics intensities prior to selecting parametric t-tests/ANOVA.
Input Variables
1 continuous numeric vector (n between 3 and 5000).
Application
Choosing between parametric and non-parametric tests shouldn't be guesswork. Shapiro-Wilk certifies whether your normalized data meets Gaussian assumptions.
#064 — D'Agostino's K-Squared Omnibus Normality Test
Statistical Definition
Combines sample skewness and kurtosis transformations into an omnibus Chi-square test of Gaussianity.
Bioinformatics Use Case
Validating normality of large-scale phenotypic distributions or residuals in genome-wide QTL mapping.
Input Variables
1 continuous numeric vector (n >= 20).
Application
In large genomic cohorts where subtle skewness or heavy tails distort linear QTL scans, D'Agostino's K-squared diagnoses exact departures from normality.
#065 — Levene's Test for Homogeneity of Variance
Statistical Definition
Tests whether k >= 2 groups have equal population variances by performing an ANOVA on absolute deviations from group means (or medians via Brown-Forsythe).
Bioinformatics Use Case
Testing differential dispersion (epigenetic/transcriptional noise) between aging and young cells, or checking ANOVA homoscedasticity.
Input Variables
k >= 2 continuous numeric vectors.
Application
Aging and cancer don't just shift average gene expression—they increase transcriptional noise. Levene's test detects differential variance across biological states.
#066 — Bartlett's Test for Equality of Variances
Statistical Definition
Parametric likelihood-ratio test evaluating whether k >= 2 normally distributed samples share identical variance.
Bioinformatics Use Case
Quality control comparing technical replicate variance across multiple mass-spectrometry runs or sequencing flowcells.
Input Variables
k >= 2 normally distributed continuous vectors.
Application
Verify instrument stability across multiple sequencing runs or assay plates with Bartlett's high-sensitivity variance equality test.
#067 — F-Test for Equality of Two Variances
Statistical Definition
Tests whether two independent normal populations have the same variance using the ratio of their sample variances.
Bioinformatics Use Case
Comparing precision (variance) of a novel diagnostic assay against an existing reference assay.
Input Variables
2 independent normal numeric vectors (x1, x2).
Application
Need to prove your streamlined assay matches or beats the precision of a legacy protocol? The variance ratio F-test quantifies relative measurement precision.
#068 — Fligner-Killeen Non-Parametric Variance Test
Statistical Definition
Distribution-free test for homogeneity of variances based on ranks of absolute deviations, highly resistant to non-normality.
Bioinformatics Use Case
Testing variance heterogeneity in skewed single-cell gene expression or microbiome relative abundances.
Input Variables
k >= 2 continuous numeric vectors.
Application
When data is heavily skewed and standard variance tests fail, Fligner-Killeen delivers rock-solid detection of dispersion shifts across cell populations.
#069 — Hartigans' Dip Test for Unimodality
Statistical Definition
Measures the maximum difference between the empirical distribution function and the closest unimodal distribution function to detect multimodality.
Bioinformatics Use Case
Identifying bimodal gene expression switches (ON/OFF states) across single cells or tumor patient subpopulations.
Input Variables
1 continuous numeric vector (x).
Application
A biomarker with two distinct peaks reveals hidden patient subtypes or cellular bistability. Hartigans' Dip Test statistically proves your biomarker splits patients into distinct subgroups.
#070 — Chi-Square Goodness-of-Fit Test
Statistical Definition
Tests whether observed categorical frequencies follow a specified theoretical probability distribution.
Bioinformatics Use Case
Testing whether genomic variant classes (missense, nonsense, synonymous) or CRISPR guide dropout counts match expected neutral ratios.
Input Variables
Vector of observed integer counts (O); vector of expected counts/probabilities (E).
Application
Validate whether observed mutation frequencies or CRISPR screen distributions conform to theoretical expectations or reflect true biological select