TABLE 6.

Summary of the contributions of deep learning studies for automated system implementation in CRISPR-Cas systems.

Author	Contributions
Cordero-Maldonado et al. (2019)	Utilized a deep learning software based on the open-source Inception v3 library for automated injections in zebrafish embryos. The system enabled high throughput genome editing by injecting CRISPR-Cas9 and DNA constructs as efficiently as an experienced experimentalist
Allen et al. (2022)	Introduced a flow-based imaging platform employing deep learning to study the DNA damage response in human hematopoietic stem and progenitor cells treated with CRISPR-Cas9 and recombinant adeno-associated virus. This system simplified the characterization and screening process of genome-editing parameters
Patino et al. (2021)	Developed an automated single-cell electroporation system integrating deep learning and computer vision strategies for gene editing tasks. They demonstrated its potential in high-throughput, precise cell manipulation applications by delivering gRNA complexes into an induced pluripotent stem cell (iPSC) line
Kanfer et al. (2021)	Proposed a novel pooled screening approach, AI-Photoswitchable Screening (AI-PS), which integrates convolutional neural networks with CRISPRi screening for subcellular phenotyping. Their proof-of-concept screen accurately identified essential factors mediating TFEB relocation