V2 SampleSheet Maker

Generate an Illumina SampleSheet CSV from JSON.

Supported Sections

Sections currently supported:

• Header
• Reads
• BCLConvert_Settings
• BCLConvert_Data

Installation

Installation from pypi

pip install v2-samplesheet-maker

Alternatively one may use the docker container ghcr.io/umccr/v2-samplesheet-maker:latest

Usage

Usage:
v2-samplesheet-maker <input-json> <output-csv>

Use - for input-json parameter to parse json from stdin,
Use - for output-json parameter to parse output samplesheet csv to stdout.

Examples

See examples/ for more info

Input JSON

{
  "header": {
    "file_format_version": 2,
    "run_name": "my-illumina-sequencing-run",
    "run_description": "A test run",
    "instrument_platform": "NovaSeq 6000",
    "instrument_type": "NovaSeq"
  },
  "reads": {
    "read_1_cycles": 151,
    "read_2_cycles": 151,
    "index_1_cycles": 10,
    "index_2_cycles": 10
  },
  "bclconvert_settings": {
    "adapter_behavior": "trim",
    "adapter_read_1": null,
    "adapter_read_2": null,
    "adapter_stringency": null,
    "barcode_mismatches_index_1": 1,
    "barcode_mismatches_index_2": 1,
    "minimum_trimmed_read_length": null,
    "minimum_adapter_overlap": 2,
    "mask_short_reads": null,
    "override_cycles": "Y151;Y10;Y8N2;Y151",
    "trim_umi": null,
    "create_fastq_for_index_reads": false,
    "no_lane_splitting": false,
    "fastq_compression_format": "gzip",
    "find_adapters_with_indels": null,
    "independent_index_collision_check": null
  },
  "bclconvert_data": [
    {
      "sample_id": "MyFirstSample",
      "lane": 1,
      "index": "AAAAAAAAAA",
      "index2": "CCCCCCCC",
      "sample_project": "SampleProject",
      "sample_name": null
    },
    {
      "sample_id": "MySecondSample",
      "lane": 1,
      "index": "GGGGGGGGGG",
      "index2": "TTTTTTTT",
      "sample_project": "SampleProject",
      "sample_name": null
    }
  ]
}

Output CSV

[Header]
FileFormatVersion,2
RunName,my-illumina-sequencing-run
RunDescription,A test run
InstrumentPlatform,NovaSeq 6000
InstrumentType,NovaSeq

[Reads]
Read1Cycles,151
Read2Cycles,151
Index1Cycles,10
Index2Cycles,10

[BCLConvert_Settings]
AdapterBehavior,trim
BarcodeMismatchesIndex1,1
BarcodeMismatchesIndex2,1
MinimumAdapterOverlap,2
OverrideCycles,Y151;Y10;Y8N2;Y151
CreateFastqForIndexReads,False
NoLaneSplitting,False
FastqCompressionFormat,gzip

[BCLConvert_Data]
Lane,Sample_ID,index,index2,Sample_Project
1,MyFirstSample,AAAAAAAAAA,CCCCCCCC,SampleProject
1,MySecondSample,GGGGGGGGGG,TTTTTTTT,SampleProject

Testing your outputs

The v2-samplesheet-maker will validate that your parameters are the correct type, the correct name AND within the appropriate range. It will NOT however determine if your settings are appropriate or compatible with a BCLConvert run, i.e it will NOT ensure that your indexes have an appropriate hamming distance.

If you wish to validate your output samplesheet with BCLConvert, you can use the following script -

scripts/build-samplesheet-and-validate-with-bcl-convert.sh <input-json> [docker-image]. This will convert the input json to csv and then run the docker container bclconvert against your samplesheet.

You will need both jq and docker installed for this to work.

Additional Usage

Parsing a SampleSheet CSV to JSON

We also have created a parser between the csv and json, with the following usage syntax

Input SampleSheet CSV

[Header]
FileFormatVersion,2
RunName,my-illumina-sequencing-run
RunDescription,A test run
InstrumentPlatform,NovaSeq 6000
InstrumentType,NovaSeq

[Reads]
Read1Cycles,151
Read2Cycles,151
Index1Cycles,10
Index2Cycles,10

[BCLConvert_Settings]
AdapterBehavior,trim
BarcodeMismatchesIndex1,1
BarcodeMismatchesIndex2,1
MinimumAdapterOverlap,2
OverrideCycles,Y151;Y10;Y8N2;Y151
CreateFastqForIndexReads,False
NoLaneSplitting,False
FastqCompressionFormat,gzip

[BCLConvert_Data]
Lane,Sample_ID,index,index2,Sample_Project
1,MyFirstSample,AAAAAAAAAA,CCCCCCCC,SampleProject
1,MySecondSample,GGGGGGGGGG,TTTTTTTT,SampleProject

# Initialise new index variable
NEW_I7_INDEX="TTTTTTTTTT"
SAMPLE_ID="MySecondSample"

# Convert the SampleSheet to JSON
# Update the index for the sample "MySecondSample" in the bclconvert_data section
# Write out the new samplesheet to csv
v2-samplesheet-to-json SampleSheet.csv - | \
jq --raw-output \
  --arg sample_id "${SAMPLE_ID}" \
  --arg replacement_index "${NEW_I7_INDEX}" \
  '
    .bclconvert_data=( 
      .bclconvert_data |
      map(
        if .sample_id == $sample_id then
          .index = $replacement_index
        else
          .
        end
      )
    )
  ' | \
v2-samplesheet-maker - SampleSheet.updated_index.csv 

Gives us the following samplesheet as an output

[Header]
FileFormatVersion,2
RunName,my-illumina-sequencing-run
RunDescription,A test run
InstrumentPlatform,NovaSeq 6000
InstrumentType,NovaSeq

[Reads]
Read1Cycles,151
Read2Cycles,151
Index1Cycles,10
Index2Cycles,10

[BCLConvert_Settings]
AdapterBehavior,trim
BarcodeMismatchesIndex1,1
BarcodeMismatchesIndex2,1
MinimumAdapterOverlap,2
OverrideCycles,Y151;Y10;Y8N2;Y151
CreateFastqForIndexReads,False
NoLaneSplitting,False
FastqCompressionFormat,gzip

[BCLConvert_Data]
Lane,Sample_ID,index,index2,Sample_Project
1,MyFirstSample,AAAAAAAAAA,CCCCCCCC,SampleProject
1,MySecondSample,TTTTTTTTTT,TTTTTTTT,SampleProject

If we run a diff on our original and updated samplesheet, we can see that the index for "MySecondSample" has been updated.

diff SampleSheet.csv SampleSheet.updated_index.csv

Yields

27c27
< 1,MySecondSample,GGGGGGGGGG,TTTTTTTT,SampleProject
---
> 1,MySecondSample,TTTTTTTTTT,TTTTTTTT,SampleProject

Manipulating a samplesheet by parsing through JSON

For this component, you will need some basic knowledge of the jq command line tool.

In this example we are given a standard samplesheet but want to edit the adapters for one of the samples.

Whilst we could do this with grep, sed or awk, it's a little hacky, instead we can

1. Parse the samplesheet to json to stdout
2. Use jq to update the samplesheet
3. Write out the updated samplesheet to csv

v2-samplesheet-to-json 

Contributing

Is there a missing section you'd like to see?

You can generate your own sections by

1. Forking this repository.
2. Adding the section pydantic model to the section models module
3. Extending the SampleSheel pydantic model in the samplesheet models module
4. Generating a class for your section in the section class module
5. Extending the SampleSheet class to write out your section in the samplesheet class module
6. Add a test for your section in the test_sections module
7. Generating a PR back to this repository! ( Please squash your commits :) )

Release Cycles

We try and update this repository for every new Dragen Release (which coincides with a BCLConvert release) and tag accordingly.