In maialab/mskcc.oncotree: Interface to the 'OncoTree' API

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

mskcc.oncotree

The goal of {mskcc.oncotree} is to facilitate access to the OncoTree API.

The OncoTree is an open-source ontology that was developed at Memorial Sloan Kettering Cancer Center (MSK) for standardizing cancer type diagnosis from a clinical perspective by assigning each diagnosis a unique OncoTree code.

Currently, the functionality provided is the retrieval of tumor types and the mapping of tumor type codes across a few tumor type classification systems.

Installation

Install {mskcc.oncotree} from CRAN:

install.packages("mskcc.oncotree")

You can install the development version of {mskcc.oncotree} with:

# install.packages("remotes")
remotes::install_github("maialab/mskcc.oncotree")

OncoTree tumor types

Get the tumor types defined by OncoTree in their latest release:

library(mskcc.oncotree)

(tumor_types <- get_tumor_types())

The mapping to the corresponding tumor type identifier(s) in the Unified Medical Language System (UMLS) and in the National Cancer Institute (NCI) Thesaurus classification systems is provided in columns umls_code and nci_code.

tumor_types[c('oncotree_code', 'umls_code', 'nci_code')]

umls_code and nci_code are list-columns because the mapping can be one-to-many.

If you want, you may convert these columns to type character by unnesting with tidyr::unnest(). This will have the side effect of creating more than one row per tumor type for cases where the mapping is one-to-many.

tumor_types[c('oncotree_code', 'umls_code')] %>%
  tidyr::unnest(cols = 'umls_code', keep_empty = TRUE)

OncoTree versions

The data provided by OncoTree is versioned. You may list the released versions so far with get_versions():

get_versions()

The function get_tumor_types() accepts a release version, allowing to get data for past versions of OncoTree:

get_tumor_types(oncotree_version = 'oncotree_2019_08_01')

Mapping of tumor type codes

UMLS and NCIt

OncoTree provides a few mappings between their identifiers (codes) and other tumor classification systems. Currently, through their web API, only mappings between OncoTree codes and Unified Medical Language System (UMLS) and National Cancer Institute (NCI) Thesaurus (NCIt) codes are provided.

From {mskcc.oncotree} we provide four functions:

| Function | Description | | -------------------- | -------------------------- | | oncotree_to_nci() | Map OncoTree to NCIt codes | | nci_to_oncotree() | Map NCIt to OncoTree codes | | oncotree_to_umls() | Map OncoTree to UMLS codes | | umls_to_oncotree() | Map UMLS to OncoTree codes |

From and to NCIt

oncotree_to_nci(c('MMB', 'PAOS'), expand = TRUE)

nci_to_oncotree(c('C3706', 'C8969'), expand = TRUE)

From and to UMLS

oncotree_to_umls(c('MMB', 'PAOS'), expand = TRUE)

umls_to_oncotree(c('C0205833', 'C0206642', 'C1334708'), expand = TRUE)

Other mappings

Besides UMLS and NCIt, you may also map OncoTree codes to ICD-O topography and morphology, and HemeOnc codes. This functionality is based on the file ontology_mappings.txt. provided by OncoTree at their GitHub repository.

In OncoTree's discussion forum the developers have warned that this data can't be expected to be neither accurate nor complete. However, it might still be useful for some users, and hence we also provide access to these mappings with the function map_ontology_code().

Here are some usage examples map_ontology_code():

# Simple example
map_ontology_code(code = 'MMB', from = 'oncotree_code', to = 'nci_code')

# Omit the `code` argument to get all possible mappings. Note that
# one-to-many mappings will generate more than one row per `from` code.
map_ontology_code(from = 'oncotree_code', to = 'nci_code')

# Some mappings are one-to-many, e.g. "SRCCR", which means repeated rows for
# the same input code.
map_ontology_code(code = 'SRCCR', from = 'oncotree_code', to = 'nci_code')

# Using the `collapse` argument to "collapse" one-to-many mappings makes sure
# that the output has as many rows as the `from` vector.
map_ontology_code(code = 'SRCCR',
                  from = 'oncotree_code',
                  to = 'nci_code',
                  collapse = toString)

map_ontology_code(code = 'SRCCR',
                  from = 'oncotree_code',
                  to = 'nci_code',
                  collapse = list)

map_ontology_code(
  code = 'SRCCR',
  from = 'oncotree_code',
  to = 'nci_code',
  collapse = \(x) paste(x, collapse = ' ')
)

# `map_ontology_code()` is vectorized over `code`
map_ontology_code(code = c('AASTR', 'MDEP'), from = 'oncotree_code', to = 'nci_code')

# Map from ICDO topography to ICDO morphology codes
map_ontology_code(code = 'C72.9', from = 'icdo_topography_code', to = 'icdo_morphology_code')

maialab/mskcc.oncotree documentation built on Oct. 15, 2022, 12:01 p.m.