Fig. 2

Schematic representation of Cruxome workflow. Typical Cruxome workflow contains four steps (input, format check, variants annotation and interpretation, and reporting and knowledge base), and time consumption to perform each step is indicated. After uploading a VCF file, Cruxome executes the VCF file format check in “Format check” step. Sequential “Variants annotation and interpretation” step is then launched to annotate and interpret the variants using various tools and databases. By using Natural Language Processing Algorithm NER, the hot gene panel and multiple databases, Cruxome performs integrated analyses that interpret and score variants according to ACMG guidelines. The report of the analysis is then exported in a PDF or Word format, and the personal knowledge base is automatically updated to store all the interpretation information. From sample input to report, the whole process takes approximately 30 min. OMIM: Online Mendelian Inheritance in Man, a comprehensive, authoritative compendium of human genes and genetic phenotypes; HPO: Human Phenotype Ontology, provides standardized vocabulary of phenotypic abnormalities encountered in human disease; ClinVar: a publicly available database that aggregates information about sequence variation and its relationship to human health; HGMD: Human Gene Mutation Database, represents all known (published) gene lesions responsible for human inherited disease; 1 K Genome: data from 1000 Genomes Project; ExAC: Exome Aggregation Consortium; gnomeAD: Genome Aggregation Database; In-House: genomic database of Berry Genomics Co., Ltd.; CADD: Combined Annotation Dependent Depletion, integrates multiple annotations into one metric; GERP++: Genomic Evolutionary Rate Profiling; SpliceAI: A deep learning-based tool to identify splice variants; PolyPhen-2: predicts possible impact of an amino acid substitution on the structure and function of a human protein