ADRnormit

ADRnormit is a computational tool designed for the normalization of single-cell RNA sequencing (scRNA-seq) data. Its primary function is to address batch effects and other technical variations that can arise during scRNA-seq experiments, ensuring that biological variability is more accurately represented. The method utilizes a non-linear dimensionality reduction approach, specifically inspired by techniques like t-SNE, to learn a low-dimensional embedding of the data. Within this embedding space, ADRnormit aims to align cells from different batches or experimental conditions.

The core principle behind ADRnormit is to identify and correct for systematic biases without removing genuine biological differences. It achieves this by modeling the distribution of cells in the latent space and then transforming this distribution to minimize the influence of unwanted technical factors. This approach allows for more robust downstream analyses, such as cell clustering, differential gene expression analysis, and trajectory inference, by providing a cleaner and more comparable dataset. ADRnormit is implemented as an R package, making it accessible to researchers familiar with the R programming environment for bioinformatics. Its development aims to improve the reliability and reproducibility of scRNA-seq studies.