README.md
In MSKCC-Epi-Bio/gnomeR: Wrangle and analyze IMPACT and TCGA mutation data

gnomeR

You can install the development version of gnomeR from GitHub with:

# install.packages("devtools")
devtools::install_github("MSKCC-Epi-Bio/gnomeR")

Along with its companion package for cbioPortal data download:

devtools::install_github("karissawhiting/cbioportalR")

the gnomeR package provides a consistent framework for genetic data processing, visualization and analysis. This is primarily targeted to IMPACT datasets but can also be applied to any genomic data provided by cBioPortal. With {gnomeR} and {cbioportalR} you can:

Download and gather data from CbioPortal - Pull from cBioPortal data base by study ID or sample ID.
OncoKB annotate data (coming soon) - Filter genomic data for known oncogenic alterations.
Process genomic data - Process retrieved mutation/maf, fusions, copy-number alteration, and segmentation data (when available) into an analysis-ready formats.
Visualize processed data - Create summary plots from processed data.
Analyzing processed data- Analyze associations between genomic variables and clinical variables or outcomes.

{gnomeR} works with any genomic data that follows cBioPortal guidelines for mutation, CNA, or fusion data file formats.

If you wish to pull the data directly from cBioPortal, see how to get set up with credentials with the {cbioportalR} package.

The below examples uses the data sets mutatations, sv, cna which were pulled from cBioPortal and are included in the package as example data sets. We will sample 100 samples for examples:

set.seed(123)

mut <- gnomeR::mutations
cna <- gnomeR::cna
sv <- gnomeR::sv

un <-  unique(mut$sampleId)
sample_patients <- sample(un, size = 50, replace = FALSE)

The main data processing function is create_gene_binary() which takes mutation, CNA and fusion files as input, and outputs a binary matrix of N rows (number of samples) by M genes included in the data set. We can specify which patients are included which will force all patients in resulting dataframe, even if they have no alterations.

gen_dat <- create_gene_binary(samples = sample_patients,
                         mutation = mut,
                         fusion = sv,
                         cna = cna)

head(gen_dat[, 1:6])
#> # A tibble: 6 × 6
#>   sample_id           ALK   APC    AR  ARAF   ATM
#>   <chr>             <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 P-0004508-T01-IM5     1     0     0     0     0
#> 2 P-0005806-T01-IM5     0     1     0     0     0
#> 3 P-0007006-T01-IM5     0     1     0     0     0
#> 4 P-0008682-T01-IM5     0     1     0     0     0
#> 5 P-0001297-T01-IM3     0     0     1     0     0
#> 6 P-0007538-T01-IM5     0     0     0     1     0

By default, mutations, CNA and fusions will be returned in separate columns. You can combine these at the gene level using the following:

by_gene <- gen_dat %>% 
  summarize_by_gene()

head(by_gene[,1:6])
#>           sample_id ALK ARAF BLM CD79B CSF1R
#> 1 P-0004508-T01-IM5   1    0   0     0     0
#> 2 P-0005806-T01-IM5   0    0   0     0     0
#> 3 P-0007006-T01-IM5   0    0   0     0     0
#> 4 P-0008682-T01-IM5   0    0   0     0     0
#> 5 P-0001297-T01-IM3   0    0   0     0     0
#> 6 P-0007538-T01-IM5   0    1   0     0     1

You can visualize your processed and raw alteration data sets using {gnomeR}’s many data visualization functions.

Quickly visualize mutation characteristics with ggvarclass(), ggvartype(), ggsnvclass(), ggsamplevar(), ggtopgenes(), gggenecor(), and ggcomut().

ggvarclass(mutation = mut)

You can tabulate summarize your genomic data frame using the tbl_genomic() function, a wrapper for gtsummary::tbl_summary().


gen_dat <- gen_dat %>%
  dplyr::mutate(trt_status = sample(x = c("pre-trt", "post-trt"),
       size = nrow(gen_dat), replace = TRUE))


gene_tbl_trt <-  gen_dat %>%
  subset_by_frequency(t = .1, other_vars = trt_status) %>%
  tbl_genomic(by = trt_status) %>%
  gtsummary::add_p()

Additionally, you can analyze custom pathways, or a set of default gene pathways using add_pathways():


path_by_trt <- gen_dat %>%
  add_pathways() %>%
  select(sample_id, trt_status, contains("pathway_")) %>%
  tbl_genomic(by = trt_status) %>%
  gtsummary::add_p()