In MAFDash: Create an HTML Dashboard to Visualize Data from MAF File

Variant Table

mydf <- generateVariantTable(maf)
column_labels <- setNames(colnames(mydf), make.names(colnames(mydf)))
colnames(mydf) <- names(column_labels)

n_samples <- length(unique(mydf$Sample.ID))
extra_data <- mydf %>% dplyr::group_by(Hugo.Symbol) %>% dplyr::mutate(Gene.Altered.in.Cohort.frac=length(unique(Sample.ID))/n_samples)
extra_cols <- c("Gene.Altered.in.Cohort.frac")

if ("Protein.Change" %in% colnames(extra_data)) {
  extra_data <- extra_data %>% dplyr::group_by(Hugo.Symbol, Protein.Change) %>%
                               dplyr::mutate(Variant.in.Cohort.frac=length(unique(Sample.ID))/n_samples)
  extra_cols <- c(extra_cols,"Variant.in.Cohort.frac")
}

mydf <- cbind(mydf[,1:2], extra_data[,extra_cols], mydf[,3:ncol(mydf)])



normdb_cols <- intersect(c("gnomAD.Frequency","ExAC.Frequency","X1000Genomes.Frequency"),colnames(mydf))
columns_to_round <- c("Gene.Altered.in.Cohort.frac","Variant.in.Cohort.frac", "Tumor.Alt.Frequency","Tumor.Depth",
                      normdb_cols)
columns_to_round <- columns_to_round[columns_to_round %in% colnames(mydf)]
for (colname in columns_to_round) {
  mydf[,colname] <- as.numeric(as.character(mydf[,colname]))
  mydf[,colname] <- ifelse(is.na(mydf[,colname]), 0, mydf[,colname])
  mydf[,colname] <- round(mydf[,colname],3)
}

# mydf[,"Tumor.Depth"] <-  as.numeric(as.character(mydf[,"Tumor.Depth"]))

mydf <- dplyr::mutate(mydf, mean.freq.normals = round(rowMeans(dplyr::select(mydf, dplyr::all_of(normdb_cols)), na.rm = TRUE),3))


mycounts <- mydf %>% dplyr::group_by(Sample.ID) %>% dplyr::summarise(num_mut=dplyr::n(), .groups="drop")
mydf$Sample.ID <- factor(as.character(mydf$Sample.ID), levels=unique(as.character(mycounts$Sample.ID)[order(mycounts$num_mut,decreasing = T)]))

Inputs {.sidebar data-width=300}

# Wrap data frame in SharedData
variant_sd <- SharedData$new(mydf)

# Create filter inputs
crosstalk_filters <- c("Sample ID"="Sample.ID",
                       "Gene mutated in cohort (frequency)"="Gene.Altered.in.Cohort.frac",
                       "Gene Symbol"="Hugo.Symbol",
                       "Protein change"="Protein.Change",
                       "Variant found in cohort (frequency)"="Variant.in.Cohort.frac",
                       "Total depth (count) at site"="Tumor.Depth",
                       "Frequency of mutated allele at site"="Tumor.Alt.Frequency",
                       "Type of variant (SNP, INDEL)"="Variant.Type",
                       # "Mean frequency in gnomAD, ExAC, and 1000 Genomes"="mean.freq.normals",
                       "gnomAD frequency"="gnomAD.Frequency",
                       "ExAC frequency"="ExAC.Frequency",
                       "1000Genomes frequency"="X1000Genomes.Frequency"
                       )

filters_df <- data.frame(sd_colname=crosstalk_filters,
                         filter_label=names(crosstalk_filters),
                         html_id=paste0("filter_id",1:length(crosstalk_filters)),
                         stringsAsFactors = F)
filters_df <- filters_df[filters_df$sd_colname %in% colnames(mydf),]

# print(unlist(lapply(mydf, class)))

rchunk_text <- c("``` {r}","```")
# browser()
filter_header_text <- "h4('Filter data')"
cohort_filter_text <- rep("",nrow(filters_df))
for (i in 1:nrow(filters_df)) {
  if (is.numeric(mydf[,filters_df$sd_colname[i]])) {
    cohort_filter_text[i] <- paste0("filter_slider(\"",filters_df$html_id[i],"\",\"",filters_df$filter_label[i],"\", variant_sd, column=~",filters_df$sd_colname[i],")")
  } else {
    cohort_filter_text[i] <- paste0("filter_select(\"",filters_df$html_id[i],"\",\"",filters_df$filter_label[i],"\", variant_sd, group=~",filters_df$sd_colname[i],")")
  }
}
# cohort_knit_text <- knitr::knit_expand(text=cohort_filter_text)
cohort_knit_text <- c(rchunk_text[1], filter_header_text, cohort_filter_text, rchunk_text[2])

r knitr::knit(text = paste(cohort_knit_text, collapse = '\n'))

# paste(knitr::knit(text = paste(cohort_knit_text, collapse = '\n')))

crosstalk_filters <- c("Sample ID"="Sample.ID",
                       "Gene mutated in cohort (frequency)"="Gene.Altered.in.Cohort.frac",
                       "Gene Symbol"="Hugo.Symbol",
                       "Protein change"="Protein.Change",
                       "Variant found in cohort (frequency)"="Variant.in.Cohort.frac",
                       "Total depth (count) at site"="Tumor.Depth",
                       "Frequency of mutated allele at site"="Tumor.Alt.Frequency",
                       "Type of variant (SNP, INDEL)"="Variant.Type",
                       "Mean frequency in gnomAD, ExAC, and 1000 Genomes"="mean.freq.normals",
                       "gnomAD frequency"="gnomAD.Frequency",
                       "ExAC frequency"="ExAC.Frequency",
                       "1000Genomes frequency"="X1000Genomes.Frequency"
                       )
# filter_select("sample_filter", "Sample ID", variant_sd, group=~Sample.ID)
# filter_select("sample_filter", "Sample ID", variant_sd, group=~Sample.ID)
# filter_slider("fracgene_filter", "Gene mutated in cohort (frequency)", variant_sd, column=~Gene.Altered.in.Cohort.frac, step=0.01, width=250)
# br()
# h4("Filter data by variant values")
# filter_select("gene_filter", "Gene Symbol", variant_sd, group=~Hugo.Symbol)
# filter_select("protchange_filter", "Protein change", variant_sd, group=~Protein.Change)
# filter_slider("fracvar_filter", "Variant found in cohort (frequency)", variant_sd, column=~Variant.in.Cohort.frac, step=0.01, width=250)
# filter_slider("tumdepth_filter", "Total depth (count) at site", variant_sd, column=~Tumor.Depth, step=5, width=250)
# filter_slider("tumfreq_filter", "Frequency of mutated allele at site", variant_sd, column=~Tumor.Alt.Frequency, step=0.01, width=250)
# filter_select("vartype_filter", "Type of variant (SNP, INDEL)", variant_sd, group=~Variant.Type)
# br()
# h4("Filter common variants in normal databases")
# filter_slider("normal_mean_freq_filter", "Mean frequency in gnomAD, ExAC, and 1000 Genomes", variant_sd, column=~mean.freq.normals, step=0.01, width=250)
# filter_slider("gnomad_filter", "gnomAD frequency", variant_sd, column=~gnomAD.Frequency, step=0.01, width=250)
# filter_slider("exac_filter", "ExAC frequency", variant_sd, column=~ExAC.Frequency, step=0.01, width=250)
# filter_slider("kgenomes_filter", "1000Genomes frequency", variant_sd, column=~X1000Genomes.Frequency, step=0.01, width=250)
br()

Row {data-height=500}

Mutation Info Table

DT::datatable(variant_sd, extensions=c("Scroller","FixedColumns","Buttons","ColReorder"), style="bootstrap", class=c("compact","display"), width="90%",
          fillContainer=F,
          escape = F,
          selection = "none",
          rownames=F,
          colnames=tools::toTitleCase(gsub("_"," ", colnames(mydf))),
          options=list(dom = 'BSRlfrtip',
                       deferRender=TRUE,
                       scrollX=T,
                       scrollY=450,
                       scroller=T,
                       buttons =list('copy', 'print', list(
                                    extend = 'collection',
                                    buttons = c('csv', 'excel', 'pdf'),
                                    text = 'Download'
                                  )),
                       colReorder=TRUE,
                       fixedColumns = list(leftColumns = 4)
                       )
          )