Detailed data model

Principles

Heliotrope uses a document-oriented database, MongoDB, because of its schema-less nature. This allows it to incorporate future data without affecting existing data structures. There are a number of primary collections:

• genes - data associated with genes
• variants - data associated with variants
• annotations - both user- and system-supplied annotations, keyed to either a variant or a gene
• tags - classifications for variant types, used for annotations
• statistics - precomputed statistical results slow to calculate dynamically, e.g., the primary gene frequencies
• variantRecords - sample-level variant occurrences, used to generate frequency information (mainly from COSMIC)

Genes

See the [gene data model page](Gene Data Model)

Variants

Positions

This is stored in sections.positions.data, and is a list of objects with the following fields:

• start - integer - the start position on the chromosome
• stop - integer - the stop position on the chromosome
• chromosome - string - the chromosome name, i.e., 1-22, X, Y, ...
• strand - integer - the strand, either +1 or -1
• cdsPosition - integer - the start of the variant as a coding position
• codon - integer - the start of the variant as a codon number
• exon - integer - the exon number
• HGVSc - string - the variant in HGVS nomenclature, DNA system
• HGVSp - string - the variant in HGVS nomenclature, amino-acid system
• HGVSpr - string - the variant in HGVS nomenclature, amino-acid system, using letters instead of codes
• referenceAllele - string - the reference allele for a mutation
• variantAllele - string - the variant allele for a mutation
• consequence - string - the consequence type, e.g., synonymous_variant
• gene - Ensembl identifier for the gene
• transcript - Ensembl identifier for the transcript for this position
• sift - object - an object with two keys, a level and a score, for SIFT prediction
• polyphen - object - an object with two keys, a level and a score, for Polyphen prediction
• significance - string - from dbSnp, e.g., pathogenic

Notes:

1. Only positions for the canonical transcript are stored.