Additional plots and analysis for neoepitope elimination analysis

Setup

library(ithi.utils)
load_base_libs()

library(methods)

library(ithi.meta)
library(ithi.figures)
library(ithi.utils)
library(ithi.seq)
library(ithi.clones)
library(ithi.supp)

ihc_table_path <- snakemake@input$ihc_table
neoediting_outdir <- snakemake@input$neoediting_outdir
snv_cluster_files <- snakemake@input$snv_cluster_files
clone_tree_file <- snakemake@input$clone_tree_file
clone_branch_length_file <- snakemake@input$clone_branch_length_file
clone_prevalence_file <- snakemake@input$clone_prevalence_file

db_path <- snakemake@params$db

annotation_colours <- ithi.figures::get_annotation_colours()

ihc_table <- fread(ihc_table_path)

tree_branch_data <- read_clone_tree_data(clone_tree_file, clone_branch_length_file, 
    clone_prevalence_file, db_path)

neoediting_res <- supp_neoediting(neoediting_outdir, ihc_table, db_path, tree_branch_data, 
    wtfilter = TRUE, full_epitopes = FALSE, snv_cluster_files = snv_cluster_files)

Analysis

This comment just seems to want some plots and an additional supplementary table. Since we are short on figure space, we’ll probably have to give this a part of a main figure – shared with the neoantigen validation experiments perhaps also the LOHHLA stuff.

grid.newpage()
grid.draw(neoediting_res$fig[[1]])

These are the patients showing high variability in the x-axis. While patient 17 looks to be going in the opposite direction as everything else – do note that the values on the y-axis are very close to eahc other on that plot (ranging from 1 to approx. 1.1, unlike the other patients).

grid.newpage()
grid.draw(neoediting_res$fig[[2]])

These are the patients with very low variability in the x-axis. These contribute very little to the statistical result since there’s virtually no absolute epithelial TIL density differences between samples.

TOOD: Check what TIL clusters these patients have.

In another file we address the potential issue that tumour cellularity may be the factor behind this – by including it into the GLM and showing that the results remain significant.

We have to provide a supplementary table with these values as well – that’s something I’ll do separately in the paper pipeline.