The paper introduces DeepVariant, a deep learning-based approach for calling genetic variants from next-generation sequencing (NGS) data. Traditional variant callers, such as GATK, rely on hand-crafted statistical models that can produce thousands of errors and missed variants. DeepVariant uses a deep convolutional neural network (CNN) to learn statistical relationships between images of read pileups around putative variant sites and ground-truth genotype calls. This approach outperforms existing tools, winning the "highest performance" award for SNPs in a FDA-administered variant calling challenge. DeepVariant generalizes across genome builds and even to other mammalian species, allowing non-human sequencing projects to benefit from human ground truth data. The model can also learn to call variants in various sequencing technologies and experimental designs, demonstrating its robustness and versatility. The paper provides detailed descriptions of the DeepVariant workflow, including candidate variant identification, image creation, and deep learning model training and inference.The paper introduces DeepVariant, a deep learning-based approach for calling genetic variants from next-generation sequencing (NGS) data. Traditional variant callers, such as GATK, rely on hand-crafted statistical models that can produce thousands of errors and missed variants. DeepVariant uses a deep convolutional neural network (CNN) to learn statistical relationships between images of read pileups around putative variant sites and ground-truth genotype calls. This approach outperforms existing tools, winning the "highest performance" award for SNPs in a FDA-administered variant calling challenge. DeepVariant generalizes across genome builds and even to other mammalian species, allowing non-human sequencing projects to benefit from human ground truth data. The model can also learn to call variants in various sequencing technologies and experimental designs, demonstrating its robustness and versatility. The paper provides detailed descriptions of the DeepVariant workflow, including candidate variant identification, image creation, and deep learning model training and inference.