Improvements in variant calling sensitivity and specificity in single-cell DNA sequencing using deep learning

Through single-cell sequencing technologies, it is now possible to interrogate thousands of cells in a single experiment for genetic variability. Single-cell DNA platforms like Tapestri are still susceptible to errors from polymerase incorporations, structure-induced template switching, PCR mediated recombination in the workflow, or DNA-damage. Errors from sequencing could propagate from cluster amplification, cycle sequencing, or image analysis. Altogether, these errors can be divided into substitutions, insertions, and deletion errors, which range from 0.5% to 2%, depending on the sequencer. This makes rare variant and minimal residual disease detection challenging. To address these challenges, deep learning models have been developed to correct the errors, reduce false-positive rates, and predict true variants.