RUBICON: A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers
)
Background: Genomic Basecalling
Basecalling is the first step in the genomics pipeline that converts noisy electrical signals to nucleotide bases (i.e., A, C, G, T). Modern basecallers use complex deep learning-based models. The accuracy and speed of basecalling have critical implications for all the steps in genome analysis.
