Skip to content

Data Schemas

Pydantic schemas defining the SILO database format for reads.

These schemas validate read data before submission to SILO, ensuring all records conform to the expected format. Sequences follow IUPAC nucleotide codes.

Output Format

The output format for LAPIS-SILO v0.8.0+ uses:

  • camelCase naming for metadata fields (e.g., readId, sampleId, batchId)
  • Flat structure with metadata at root level (no nested "metadata" object)
  • Segment format with sequence, insertions, and offset fields
  • Insertion format as "position:sequence" (e.g., "123:ACGT")

Modifying the Schema

To customize the output schema for your use case:

  1. Edit the schema file at src/sr2silo/silo_read_schema.py:
  2. Add or modify fields in the ReadMetadata class

  3. Use Pydantic Field(alias="camelCaseName") for camelCase output

  4. Update the database config at resources/silo/database_config.yaml:

  5. Ensure field names match the Pydantic aliases

  6. Run validation to verify consistency: 

    python tests/test_database_config_validation.py

This validation ensures your Pydantic schema matches the SILO database configuration.

ReadMetadata

sr2silo.silo_read_schema.ReadMetadata

Bases: BaseModel

V-Pipe SILO-specific pydantic schema for ReadMetadata JSON format. (specific to WISE / run at ETHZ)

Uses camelCase aliases for serialization to match SILO database schema, while maintaining snake_case field names for Python conventions.

NucleotideSegment

sr2silo.silo_read_schema.NucleotideSegment

Bases: BaseModel

Schema for a nucleotide genomic segment (e.g., main nucleotide sequence).

validate_nucleotide_insertions(v) classmethod

Validate that nucleotide insertions have the format 'position:sequence'.

validate_nucleotide_sequence(v) classmethod

Validate that sequence contains only valid nucleotide characters.

validate_offset(v) classmethod

Validate that offset is non-negative.

AminoAcidSegment

sr2silo.silo_read_schema.AminoAcidSegment

Bases: BaseModel

Schema for an amino acid genomic segment (e.g., gene sequences).

validate_amino_acid_insertions(v) classmethod

Validate that amino acid insertions have the format 'position:sequence'.

validate_amino_acid_sequence(v) classmethod

Validate that sequence contains only valid amino acid characters.

validate_offset(v) classmethod

Validate that offset is non-negative.

AlignedReadSchema

sr2silo.silo_read_schema.AlignedReadSchema

Bases: BaseModel

SILO-specific pydantic schema for AlignedRead JSON format.

This schema validates the new SILO format where: - readId is a required field at the root level (uses camelCase alias) - Metadata fields are at the root level (use camelCase aliases) - 'main' is a nucleotide segment with nucleotide-specific validation - Gene segments are amino acid segments with amino acid-specific validation - Unaligned sequences are prefixed with 'unaligned_'

validate_dynamic_fields()

Validate gene segments, unaligned sequences, and metadata fields.