Transparent, adaptable, and reproducible data analyses for oncology - Snakemake, Datavzrd, Vembrane and beyond

Johannes Köster

 

2026

 

University of Duisburg-Essen

  • check computational validity
  • apply same analysis to new data
  • check methodological validity
  • understand analysis

Data analysis

Reproducibility

Transparency

  • modify analysis
  • extend analysis

Adaptability

>1.6 million downloads since 2015

>3000 citations

>14 citations per week in 2024

dataset

results

dataset

dataset

dataset

dataset

dataset

Define workflows

in terms of rules

Define workflows

in terms of rules

rule mytask:
    input:
        "path/to/{dataset}.txt"
    output:
        "result/{dataset}.txt"
    script:
        "scripts/myscript.R"


rule myfiltration:
     input:
        "result/{dataset}.txt"
     output:
        "result/{dataset}.filtered.txt"
     shell:
        "mycommand {input} > {output}"


rule aggregate:
    input:
        "results/dataset1.filtered.txt",
        "results/dataset2.filtered.txt"
    output:
        "plots/myplot.pdf"
    script:
        "scripts/myplot.R"

Define workflows

in terms of rules

Define workflows

in terms of rules

rule mytask:
    input:
        "data/{sample}.txt"
    output:
        "result/{sample}.txt"
    shell:
        "some-tool {input} > {output}"

rule name

how to create output from input

define

  • input
  • output
  • log files
  • parameters
  • resources
rule mytask:
    input:
        "path/to/{dataset}.txt"
    output:
        "result/{dataset}.txt"
    script:
        "scripts/myscript.R"


rule myfiltration:
     input:
        "result/{dataset}.txt"
     output:
        "result/{dataset}.filtered.txt"
     shell:
        "mycommand {input} > {output}"


rule aggregate:
    input:
        "results/dataset1.filtered.txt",
        "results/dataset2.filtered.txt"
    output:
        "plots/myplot.pdf"
    script:
        "scripts/myplot.R"

Automatic inference of DAG of jobs

Boilerplate-free integration of scripts

rule mytask:
    input:
        "data/{sample}.txt"
    output:
        "result/{sample}.txt"
    script:
        "scripts/myscript.py"

reusable scripts:

  • Python
  • R
  • Julia
  • Rust
  • Bash
  • Xonsh
  • Hy

Reusable wrappers

rule map_reads:
    input:
        "{sample}.bam"
    output:
        "{sample}.sorted.bam"
    wrapper:
        "0.22.0/bio/samtools/sort"

reuseable wrappers from central repository

Reusable wrappers

Self-contained HTML reports

Plugins

Many more features

  • dynamic DAG rewiring
  • service jobs (providing sockets, loading databases, or ramdisks)
  • semantic helper functions for minimizing boilerplate code
  • fallible rules
  • caching of shared results across workflows
  • transparent handling of remote storage

Tables are the central entity
in data analysis

https://datavzrd.github.io

Not always a single table

oncoprint + individual variant calls

differentially expressed genes + expression matrix

https://datavzrd.github.io

Not always just a table

oncoprint + individual variant calls + 

differentially expressed genes + expression matrix +

https://datavzrd.github.io

State of the art

Individual tables (tsv, excel) and plots:

  • easy to publish
  • limited interactivity
  • no jumping between corresponding items

 

Web applications (custom, shiny, ...):

  • running server (or local installation)
  • implementation overhead
  • long-term maintenance is challenging

https://datavzrd.github.io

The problem

Input:

  • set of tables
  • relations between tables
  • set of rendering definitions

Output:

portable interactive visual presentation

https://datavzrd.github.io

https://datavzrd.github.io

Datavzrd

https://datavzrd.github.io

Datavzrd

https://datavzrd.github.io

Datavzrd + Snakemake

rule datavzrd:
    input:
        config="resources/{sample}.datavzrd.yaml",
        table="data/{sample}.tsv",
    output:
        report(
            directory("results/datavzrd-report/{sample}"),
            htmlindex="index.html",
        ),
    wrapper:
        "v4.6.0/utils/datavzrd"

https://datavzrd.github.io

Vembrane

harmonize

  • variant filtration
  • variant prioritization
  • variant exchange (FHIR)

https://vembrane.github.io

(not {'risk_factor','pathogenic', 'drug_response'}.isdisjoint(ANN['CLIN_SIG'])) and
(ANN['IMPACT'] in {'LOW', 'MODERATE', 'HIGH'}) and
(ANN['REVEL'] is NA or ANN['REVEL'] >= 0.5)

pathogenic/risk-factor/drug-response (vembrane filter)

relevant VAFs (vembrane filter)

(FORMAT['AF']['tumor'] >= 0.05 and (ANN['gnomADg_AF'] is NA or ANN['gnomADg_AF'] < 0.01))
desc(quantize(max(FORMAT['AF'][sample] for sample in SAMPLES), 1 / 3)), \
desc(ANN["REVEL"]), \
desc(FORMAT['AF']['tumor'])

sort by descending relevance (vembrane sort)

Vembrane

Adaptable and unified small-variant/structural-variant/fusion calling on any scenario

Snakemake + Datavzrd + Vembrane + Varlociraptor =

reference pipeline project

samples:
  jane:
    sex: female
    somatic-effective-mutation-rate: 1e-10
  tumor:
    inheritance:
      clonal:
        from: jane
    contamination:
      by: jane
      fraction: 0.1
    somatic-effective-mutation-rate: 1e-6
  relapse:
    inheritance:
      clonal:
        from: jane
    contamination:
      by: jane
      fraction: 0.2
    somatic-effective-mutation-rate: 1e-6
expressions:
  somatic_tumor: "jane:0.0 & tumor:]0.0,1.0]"

events:
  germline: "jane:0.5 | jane:1.0"
  somatic: "jane:]0.0,0.5["
  somatic_tumor_no_increase: "$somatic_tumor & l2fc(relapse,tumor) < 1"
  somatic_tumor_increase: "$somatic_tumor & l2fc(relapse,tumor) >= 1"
  somatic_relapse: "jane:0.0 & tumor:0.0 & relapse:]0.0,1.0]"

https://varlociraptor.github.io

varlociraptor

Dual use in research and clinics

Conclusion

Snakemake + Datavzrd

generic, transparent, human readable, adaptable, portable, and scalable data analysis and visualization

 

Vembrane

Unified DSL for filtering, sorting and transforming genomic variants/fusions/CNVs

 

Varlociraptor + Pipeline

unified, adaptable small/structural variant calling for any scenario