format_gistic.Rmd

---
title: "Process Gistic Files"
author: "Ted Laderas"
date: "8/15/2019"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Transforming the Gistic Data

This is a simple script that transforms the GISTIC calls for the HNSCC data (the `all_thresholded.by_genes.txt") into a long format more appropriate to introduce into the mutation workflow.

```{r}
library(data.table)
library(dtplyr)
library(dplyr)
library(here)


gistic_file <- data.table::fread(here("data/gdac.broadinstitute.org_HNSC-TP.CopyNumber_Gistic2.Level_4.2012122100.0.0/all_thresholded.by_genes.txt"))

gistic_long <- gistic_file %>% tidyr::gather("tcga_id", "cnv_call", - `Gene Symbol`, - `Locus ID`, -Cytoband) 


gistic_long <- gistic_long %>% mutate(tcga_id = substr(tcga_id, 1, 12)) %>% filter(cnv_call != 0)

#filter only to -2 and +2 calls
gistic_long_calls <- gistic_long %>% 
  filter(cnv_call !=0 & (cnv_call == -2 | cnv_call == 2)) %>%
  mutate(Variant_Classification = "CNV", 
         Variant_Type = case_when(cnv_call == -2 ~ "DEL",
                   cnv_call == 2 ~ "AMP"
  ))

saveRDS(gistic_long_calls, file=here("data/gistic_long_calls.rds"))

```


```{r}
head(gistic_long_calls)
```

```{r}
library(data.table)
library(dplyr)
gistic_long_calls <- readRDS(here::here("data/gistic_long_calls.rds"))

##these are the needed columns
cols <- c("Hugo_Symbol", "Chrom","Ncbi_Build", 
          "Start_Position", "Reference_Allele", 
          "Tumor_Sample_Barcode",  "Tumor_Seq_Allele1", 
          "Tumor_Seq_Allele2", "Variant_Classification", 
          "Variant_Type", "Dbsnp_Rs","Dbsnp_Val_Status")

tcga_samp <- fread(here::here("data/TCGA.HNSC.HPV-positive.maf"))

head(tcga_samp)

new_gistic_long <- gistic_long_calls %>% mutate(Hugo_Symbol=`Gene Symbol`, 
                                                Chrom = Cytoband, 
                             Tumor_Sample_Barcode = tcga_id, 
                             Start_Position = NA, Reference_Allele = NA,
                             Ncbi_Build = "GRCh38", Tumor_Seq_Allele1=NA, 
                             Tumor_Seq_Allele2 = NA, 
                             Dbsnp_Rs = NA, Dbsnp_Val_Status = NA, Sequencer=NA
                             )

write.table(new_gistic_long,file = here::here("data/gistic_long_file.maf"), sep="\t", row.names = FALSE)

#show that this dummied data passes dataCleaningAndCheck
try(packageDir:::dataCleaningAndCheck(head(new_gistic_long)))

#show that it passes addPidColumn
try(packageDir:::addPidColumn(head(new_gistic_long)))

#
try(packageDir:::FilterDuplicates(new_gistic_long))
```