geneOncoX.Rmd

---
title: "Getting started"
output: 
  rmarkdown::html_vignette:
    toc: true
vignette: >
  %\VignetteIndexEntry{Getting started}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
bibliography: '`r system.file("bibliography.bib", package = "geneOncoX")`'
nocite: |
  @Martin2019-nq, @Lever2019-xp, @Repana2019-dd, 
  @Bailey2018-os, @Sondka2018-wf, @Liu2020-ga, 
  @Martinez-Jimenez2020-qx, @Wood2001-nk
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

```{r setup, echo = F}
library(geneOncoX)
```

<br><br>

# Installation

**IMPORTANT NOTE**: _geneOncoX_ requires that you have R version 4.1 or higher installed

```{r install, echo = T, eval = F}

if (!("remotes" %in% installed.packages())) {
  install.packages("remotes")
}
remotes::install_github('sigven/geneOncoX')
```

<br><br>

# Get basic gene annotations

This shows how to retrieve basic gene and cancer relevant gene annotations,
including how to retrieve tumor suppressor genes, proto-oncogenes, and 
predicted cancer driver genes.

```{r basic, echo = T}
library(geneOncoX)

## load the data
download_dir <- tempdir()
gene_basic <- get_basic(cache_dir = download_dir)

## Number of records
nrow(gene_basic$records)

## Show metadata for underlying resources
gene_basic$metadata


```

<br><br>

## Get classified tumor suppressor genes

```{r tsg, echo = T, eval = T, results = "hide"}

## Get tumor suppressor genes - as indicated from either
## Cancer Gene Census (CGC) or Network of Cancer Genes (NCG)
## - show literature support from CancerMine

tsg <- gene_basic$records |>
  dplyr::filter(cgc_tsg == TRUE |
                  ncg_tsg == TRUE) |>
  dplyr::select(symbol, entrezgene, name, gene_biotype,
                cgc_tsg, ncg_tsg, cancermine_cit_links_tsg, 
                ncbi_function_summary) |>
  dplyr::rename(support_cancermine = cancermine_cit_links_tsg)

## Make as datatable
tsg_table <- DT::datatable(
  tsg, 
  escape = FALSE,
  extensions = c("Buttons", "Responsive"), 
  width = "100%",
  options = list(
    buttons = c("csv", "excel"), dom = "Bfrtip"))

```

<br>

```{r tsg_dt, echo = F, eval = T}
tsg_table
```

<br><br>

## Get classified proto-oncogenes

```{r oncogene, echo = T, eval = T, results = "hide"}

## Get proto-oncogenes - as indicated from either
## Cancer Gene Census (CGC) or Network of Cancer Genes (NCG),
## - show literature support from CancerMine
oncogene <- gene_basic$records |>
  dplyr::filter(cgc_oncogene == TRUE |
                  ncg_oncogene == TRUE) |>
  dplyr::select(symbol, entrezgene, 
                name, gene_biotype,
                cgc_tsg, ncg_tsg, 
                cancermine_cit_links_oncogene, 
                ncbi_function_summary) |>
  dplyr::rename(support_cancermine = 
                  cancermine_cit_links_oncogene)

## Make as datatable
oncogene_table <- DT::datatable(
  oncogene, 
  escape = FALSE,
  extensions = c("Buttons", "Responsive"), 
  width = "100%",
  options = list(
    buttons = c("csv", "excel"), 
    dom = "Bfrtip"))

```

<br>

```{r oncogene_table, echo = F, eval = T}
oncogene_table
```

<br><br>

## Get predicted cancer driver genes

```{r cancer_driver, echo = T, eval = T, results = "hide"}

## Get predicted cancer driver genes - as indicated from either
## - Cancer Gene Census (CGC) - tier1/tier2
## - Network of Cancer Genes (NCG) - canonical drivers
## - IntOGen mutational driver catalogue
## - TCGA's PanCancer driver prediction (Bailey et al., Cell, 2018)
##
##  Rank hits by how many sources that contribute to classification
##
cancer_driver <- gene_basic$records |>
  dplyr::filter(cgc_driver_tier1 == TRUE |
                cgc_driver_tier2 == TRUE |
                intogen_driver == TRUE |
                ncg_driver == TRUE |
                tcga_driver == TRUE) |>
  dplyr::mutate(driver_score = 0) |>
  dplyr::mutate(driver_score = dplyr::if_else(
    cgc_driver_tier1 == T, 
    driver_score + 2,
    as.numeric(driver_score))) |>
  dplyr::mutate(driver_score = dplyr::if_else(
    cgc_driver_tier2 == T, 
    driver_score + 1,
    as.numeric(driver_score))) |>
  dplyr::mutate(driver_score = dplyr::if_else(
    intogen_driver == T, 
    driver_score + 1,
    as.numeric(driver_score))) |>
  dplyr::mutate(driver_score = dplyr::if_else(
    ncg_driver == T, 
    driver_score + 1,
    as.numeric(driver_score))) |>
  dplyr::mutate(driver_score = dplyr::if_else(
    tcga_driver == T, 
    driver_score + 1,
    as.numeric(driver_score))) |>
  dplyr::select(symbol, entrezgene, 
                name, gene_biotype,
                cgc_driver_tier1,
                cgc_driver_tier2,
                intogen_driver,
                #intogen_phenotype,
                intogen_role,
                ncg_driver,
                ncg_phenotype,
                tcga_driver,
                driver_score,
                cancermine_cit_links_driver, 
                ncbi_function_summary) |>
  dplyr::rename(support_cancermine = 
                  cancermine_cit_links_driver) |>
  dplyr::arrange(dplyr::desc(driver_score), symbol)

## Make as datatable
driver_table <- DT::datatable(
  cancer_driver, 
  escape = FALSE,
  extensions = c("Buttons", "Responsive"), 
  width = "100%",
  options = list(
    buttons = c("csv", "excel"), 
    dom = "Bfrtip"))

```

<br>

```{r driver_table, echo = F, eval = T}
driver_table
```

<br><br>

# Get cancer predisposition genes

This show how to retrieve known cancer predisposition genes, utilizing
multiple sources, including Cancer Gene Census, Genomics England PanelApp,
TCGA's pan-cancer study of germline variants, and other/user-curated entries.

```{r gene_predisposition, echo = T}

## load the data
gene_predisposition <- get_predisposition(cache_dir = download_dir)

## Number of cancer predisposition genes
nrow(gene_predisposition$records |> dplyr::filter(
  !stringr::str_detect(cpg_source, "^ACMG_SF$")
))

## Get statistics regarding how reference sources on 
## cancer predisposition genes contribute
##
## CGC - Cancer Gene Census (germline)
## PANEL_APP - N = 43 gene panels for inherited cancer conditions/
##             cancer syndromes (Genomics England PanelApp)
## CURATED_OTHER - curated/user-contributed genes
## TCGA_PANCAN_2018 - TCGA's pancancer analysis of
##                    germline variants in cancer
##                    (Huang et al., Cell, 2019)
## ACMG_SF - Secondary findings list (ACMG, v3.1)
plyr::count(gene_predisposition$records$cpg_source) |>
  dplyr::arrange(dplyr::desc(freq)) |>
  dplyr::filter(x != "ACMG_SF")


## Cancer predisposition metadata
gene_predisposition$metadata

```


<br><br>

# Get cancer gene panels

This shows how to retrieve genes from cancer gene panels
defined in Genomics England PanelApp.

```{r gene_panels, echo = T}

## load the data
gene_panels <- get_panels(cache_dir = download_dir)

## panel data for genome build grch38
panel_data <- gene_panels$records |>
  dplyr::filter(genome_build == "grch38")

## show number of genes in each panel
gene_freq <- as.data.frame(panel_data |>
  dplyr::group_by(gepa_panel_name) |>
  dplyr::summarise(n = dplyr::n()) |>
  dplyr::arrange(desc(n)) |>
  dplyr::rename(panel_name = gepa_panel_name))

gene_freq

```

<br><br>

# Get gene aliases

This shows how to retrieve ambiguous and unambiguous gene aliases (i.e. with
respect to primary gene symbols).

```{r gene_alias, echo = T}

## load the data
gene_alias <- get_alias(cache_dir = download_dir)

## number of gene synonyms that are ambiguous
nrow(dplyr::filter(gene_alias$records, ambiguous == TRUE))

## show structure of alias records
head(gene_alias$records)

```


<br><br>

# Get GENCODE transcripts

This shows how to retrieve GENCODE transcripts for `grch37` and `grch38`.

```{r gencode, echo = T}

## load the data
gene_gencode <- get_gencode(cache_dir = download_dir)

## number of transcript records - grch37
nrow(gene_gencode$records$grch37)

## number of transcript records - grch38
nrow(gene_gencode$records$grch38)

## show colnames for transcript records
colnames(gene_gencode$records$grch38)

## show metadata for underlying resources
gene_gencode$metadata

```

<br><br>

# Get DNA repair genes

```{r dna_repair, echo = T}

## load the data
gene_dna_repair <- gene_basic$records |>
  dplyr::filter(!is.na(woods_dnarepair_class)) |>
  dplyr::select(symbol, woods_dnarepair_class,
                woods_dnarepair_activity)

## count number of genes in each class
dna_repair_class_freq <- as.data.frame(
  gene_dna_repair |>
  dplyr::group_by(woods_dnarepair_class) |>
  dplyr::summarise(n = dplyr::n()) |>
  dplyr::arrange(desc(n)) |>
  dplyr::rename(dna_repair_class = woods_dnarepair_class))

dna_repair_class_freq

```

<br><br>

# Get TSO500 genes

```{r tso500, echo = T}

## load the data
gene_tso500 <- gene_basic$records |>
  dplyr::filter(!is.na(illumina_tso500))

## number of genes covered by TSO500 panel
nrow(gene_tso500)

## types of variant types covered
illumina_tso500_variant_freq <- as.data.frame(
  gene_tso500 |>
  dplyr::group_by(illumina_tso500) |>
  dplyr::summarise(n = dplyr::n()) |>
  dplyr::arrange(desc(n)))

illumina_tso500_variant_freq

```


<br><br>

# Session Info

\vspace{5pt}
\footnotesize
```{r sessioninfo, eval = TRUE}
# set eval = FALSE if you don't want this info (useful for reproducibility) 
# to appear
sessionInfo()
```
\normalsize

<br><br>

# References