Jaccard

Jaccard computes Jaccard/Tanimoto similarity and statistical tests for binary presence–absence biological data to evaluate species co-occurrence across biogeographic units.

Key Features:

• Jaccard/Tanimoto Coefficient: Calculates the Jaccard/Tanimoto coefficient as the ratio of intersection to union for binary presence–absence datasets to quantify similarity and identify non-random species co-occurrences.
• Statistical Significance Testing: Performs hypothesis testing on Jaccard/Tanimoto coefficients with unbiased estimation of expectation and centered coefficients that account for species occurrence probabilities.
• Efficient Computation: Implements bootstrap and measurement concentration algorithms to reduce computation time while preserving p-value and false discovery rate accuracy in high-dimensional data.
• Simulation Studies: Validates estimation methods and statistical measures through comprehensive simulation studies.
• Empirical Applications: Applied to real-world datasets such as bird species across 28 islands of Vanuatu and fish species in 3347 freshwater habitats in France.

Scientific Applications:

• Ecology: Analyzes species co-occurrence and biogeographic distribution patterns using presence–absence matrices.
• Genomics: Applies Jaccard/Tanimoto-based similarity and hypothesis testing to binary genomic presence–absence data.
• Biochemistry: Supports analysis of binary biochemical datasets for co-presence and co-occurrence patterns.

Methodology:

Computational methods include calculation of the Jaccard/Tanimoto coefficient, hypothesis testing with unbiased estimation of expectation and centered coefficients accounting for occurrence probabilities, bootstrap and measurement concentration algorithms, and simulation studies for validation.