Felix Wiegand¹, David Lähnemann¹², Felix Mölder¹³, Hamdiye Uzuner¹, Adrian Prinz¹ Alexander Schramm², Johannes Köster¹²

 

¹Bioinformatics and Computational Oncology, Institute for AI in Medicine, University Hospital Essen, University of Duisburg-Essen

²German Cancer Consortium (DKTK). Partner Site University Hospital Essen

³Institute of Pathology, University Medical Center Essen

Laboratory of Molecular Oncology, West German Cancer Center, Department of Medical Oncology, University Hospital Essen

https://koesterlab.github.io

¹ https://shiny.rstudio.com

² https://bioinformaticsworkbook.org

Excel/CSV:

  • non-interactive
  • non-visual

Interactive visual document

  • browser based
  • portable, server-free
  • publishable as supplementary file
  • can handle big data

The problem

The solutions so far

²

Plots:

  • can only highlight certain aspects

Webservice/R Shiny:

  • development overhead
  • not portable (needs server or local deployment)

¹

The datavzrd solution

datasets:
  diffexp:
    path: results/diffexp.csv
    separator: "\t"

  differential_expression:
    dataset: diffexp
    render-table:
      columns:
        accession:
          display-mode: hidden
        name:
          link-to-url: "https://www.genecards.org/{name}"
        type:
          plot:
            heatmap:
              scale: ordinal
              color-scheme: category10
        F:
          plot:
            ticks:
              scale: linear

?

How to communicate derived information

Rapid text based declaration of

  • datasets (n)
  • views (mn)

Declare

  • linkouts
  • heatmaps
  • tickplots

Define custom plots

  • for individual table cells
  • for entire datasets

Jump around between corresponding elements in different views

Export views to

  • publication-ready figures
  • Excel

filter rows and display column stats

show and hide row details

Share rows

  • with mobile devices using QR codes
  • via mail using data URLs

Introducing

https://datavzrd.github.io

Datavzrd Poster

By Johannes Köster