In Bioconductor/SOUNDBoard: Statistically-informed Multi-Omics Reports for Tumor and Other Boards

Types of widgets

1. Static -- e.g., presenting static information, e.g., patient id, case number, ...

2. Interactive widgets -- JavaScript-based interactions for displaying static information that the user can manipulate, e.g., large tables, interactive plots.

3. shiny-based widgets -- Require evaluation in R, doing arbitrary 'read-only' operations that could potentially be time-consuming.

4. Linked widgets -- e.g., crosstalk -- choice of gene in table of variants updates choice of available drugs.

Generic implementations

Status: planned, beta, available

Static

• Serialized R objects, other than data.frame() or ggplot(). beta

  ```

  widget <- ... # any R object

  tbl(x, "assay") <- list( case_uid = "PATIENT_1", assay = "R object", description = "Demonstrating R object management", resource = manage(x, widget) ) ```

• tbl_board and tbl_cases. beta

  tbl(x, "board") <- list( board_uid = "Unique_Board_ID" description = "Board description" )

  tbl(x, "cases") <- list( board_uid = "Unique_Board_ID", case_uid = "Unique_Case_ID", age = 36, sex = "Male" )

Interactive

• data.frame() via DT::datatable(). beta

  ```

  assay resource is a data.frame object

  tbl(x, "assay") <- list(..., assay = "data.frameObj",

  resource = manage(x, ggplot2::ggplot(...))) # save into assays list

  tbl(x, "assay") %>% filter(case_uid == "Unique_Case_ID", assay == "data.frameObj") %>% sbreport() ```

• ggplot() via plotly::ggplotly(). beta

  ```

  assay resource is a ggplot object

  tbl(x, "assay") <- list(..., assay = "ggplotObj",

  resource = manage(x, ggplot2::ggplot(...))) # save into assays list

  tbl(x, "assay") %>% filter(case_uid == "Unique_Case_ID", assay == "ggplotObj") %>% sbreport() ```

Shiny-based

• ShinyAppWidget(). beta -specified via R code in a single file.

  ```

  myApp <- ShinyAppWidget(...) # assign shiny app as object

  tbl(x, "assay") <- list(..., assay = "myApp") # save into assays list

  tbl(x, "assay") %>% filter(case_uid == "Unique_Case_ID", assay == "myApp") %>% sbreport() ```

• ShinyAppDirWidget(). beta

• R code in ui.R / server.R files.

  ```

  myShinyDir <- ShinyAppDirWidget(resource = "path/to/app/dir")

  tbl(x, "assay") <- list(..., assay = "shinyAppDir",

  resource = manage(x, myShinyDir)) # save into assays

  tbl(x, "assay") %>% filter(case_uid == "Unique_Case_ID", assay == "shinyAppDir") ```

Linked

Specific implementations

• HeatmapWidget(). beta
• rows: genes; columns: samples
• samples clustered; bars to indicate phenotype
• interactive in the sense that hover over a cell provides row, column, value info

• implemented with ComplexHeatmap

  ```

  Make a Heatmap using the ComplexHeatmap package

  heatmap <- ComplexHeatmap::Heatmap(matrix(...))

  widget <- HeatmapWidget(heatmap) tbl(x, "assay") <- list( case_uid = "PATIENT_1", assay = "R object", description = "Demonstrating R object management", resource = manage(x, widget) ) ```

• OncoPrintWidget(). beta

• rows: genes: columns: samples
• Visualizes multiple genomic alteration events via heatmap

• implemented with ComplexHeatmap

  ```

  Use the ComplexHeatmap to make an oncoPrint graphic

  ht <- ComplexHeatmap::oncoPrint(...) tbl(x, "assay") <- list( case_uid = "TCGA-32-2615", assay = "onco", description = "Heatmap Widget Demo", resource = manage(x, ht) ) ```

• DGIdbWidget(). beta

• query DGIdb with genes, implemented with rDGIdb
• obtain a 'datatable' of 'rDGIdb::resultSummary' output ## Query the DGI servers for drug data dgi <- rDGIdb::queryDGIdb(...) tbl(x, "assay") <- list( case_uid = "TCGA-32-2615", assay = "DGIdb", description = "Druggability by Gene with 'resultSummary'", resource = manage(x, dgi) )

Bioconductor/SOUNDBoard documentation built on May 6, 2019, 7:52 a.m.