Chromatin accessibility refers to the degree to which the DNA in a cell’s nucleus is accessible and open for various cellular processes, such as transcription and DNA replication.

    scBasset, a sequence-based convolutional neural network (CNN) model, aims to predict chromatin accessibility (a binary classification task) based on single-cell ATAC-seq data.

    In this study, we present a comprehensive evaluation and improvement of scBasset's performance and interpretability.