Efficient and scalable Python implementation of ANCOM-BC and ANCOM-BC2 for omics differential abundance testing

Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC) and its successor ANCOM-BC2 are critical algorithms for differential abundance analysis. Designed for handling compositional data, they feature the ability to estimate the unknown sampling fractions and correct the bias induced by their differences. While the original R package implementing the two methods is robust for standard dataset sizes, the increasing magnitude of modern multi-omics studies presents computational challenges, often resulting in a runtime bottleneck for high-dimensional data. Here, we present native Python implementations of both ANCOM-BC and ANCOM-BC2 in the scikit-bio package, significantly accelerating the workflow and enabling their application to large-scale datasets. By leveraging vectorization with NumPy and numerical optimization in SciPy, we enhanced the efficiency of both methods without compromising their statistical accuracy. The benchmarking results revealed a dramatic increase in performance in both simulated and real datasets. We demonstrated that our Python implementations generated test results matching the original R code, while achieving 380- and 147-fold faster runtimes and 97.4% and 96.2% lower memory consumption on the largest tested dataset, respectively. The Python implementations allow for the analysis of datasets that are significantly larger than previously possible, and enable applications to alternative omics data types and complex statistical models that vary greatly by the number of samples, features, and covariates.

https://github.com/scikit-bio/scikit-bio.

N/A.