README.md

In rforbiodatascience/barcc: Illustrationg results from baracoda screnning combined with mupexi data

barcc

The packges will take pre-processed copy of raw baracoda files and mupexi files and illustrate the characteristics of the immunugenic neopeptides.

Installation and background

The packges will take pre-processed copy of raw baracoda files and mupexi files and illustrate the characteristics of the immunugenic neopeptides. The packgaes works only with murine data, The human data input is in progress.

# install.packages("devtools")
library(devtools)
devtools::install_github("rforbiodatascience/barcc")
library(barcc)

barcc is built with tidyverse and is required for running. ## required packages

library(tidyverse)
library(readxl)
library(openxlsx)
library(shiny)
library(ggplot2)
library(barcc)
library(ggrepel)
library(ggbeeswarm)
library(knitr)
library(ggseqlogo)

Load data

Baracoda files

Baracoda files from experiment must be loaded

path_ct26 <- "test_data/barracoda_output_CT26.xlsx"
path_4t1 <- "test_data//barracoda_output_4T1.xlsx"
all_ct26_barracoda_raw <- path_ct26 %>% 
    # function to import all sheets
    excel_sheets() %>% 
    # give names to each sheet
    set_names() %>% 
    # apply read_excel to each sheet, and add the number to the colum sheet
    map_df(~ read_excel(path = path_ct26, sheet = .x), .id = "sheet") 

all_4t1_barracoda_raw <- path_4t1 %>% 
    excel_sheets() %>% 
    set_names() %>% 
    map_df(~ read_excel(path = path_4t1, sheet = .x), .id = "sheet")
kable(head(all_4t1_barracoda_raw))

| sheet | barcode | sample | count.1 | input.1 | input.2 | input.3 | log_fold_change | p | -log10(p) | masked_p (p = 1 if logFC \< 0) | -log10(masked_p) | count.normalised (edgeR) | input.normalised (edgeR) | Peptide.name | HLA | Sequence | | :---------- | :------ | :--------- | ------: | ------: | ------: | ------: | ----------------: | --------: | ---------: | ------------------------------: | -----------------: | -----------------------: | -----------------------: | :----------- | :---- | :---------- | | X4T1_16.TU | A10B200 | 4T1_16-TU | 4 | 256 | 306 | 243 | -1.302920 | 0.2192027 | 0.6591540 | 1.0000000 | 0.00000 | 0.0112809 | 0.0278560 | 4T1_1 | H-2Db | KTVFNSRFL | | X4T1_16.TU | A15B201 | 4T1_16-TU | 0 | 81 | 105 | 83 | -9.088002 | 0.0115297 | 1.9381811 | 1.0000000 | 0.00000 | 0.0000000 | 0.0093040 | 4T1_10 | H-2Db | STSIGSSTPM | | X4T1_16.TU | A18B203 | 4T1_16-TU | 7 | 43 | 59 | 47 | 1.615952 | 0.0897883 | 1.0467802 | 0.0897883 | 1.04678 | 0.0364584 | 0.0118618 | 4T1_100 | H-2Kd | AYVKKFSYQSI | | X4T1_16.TU | A18B205 | 4T1_16-TU | 0 | 61 | 66 | 52 | -8.507643 | 0.0439994 | 1.3565535 | 1.0000000 | 0.00000 | 0.0000000 | 0.0142728 | 4T1_101 | H-2Kd | KYIIQAGQV | | X4T1_16.TU | A18B204 | 4T1_16-TU | 0 | 36 | 31 | 39 | -7.767225 | 0.0713457 | 1.1466323 | 1.0000000 | 0.00000 | 0.0000000 | 0.0085468 | 4T1_102 | H-2Kd | EYLTAKNMKL | | X4T1_16.TU | A18B206 | 4T1_16-TU | 2 | 79 | 121 | 115 | -1.269234 | 0.3793029 | 0.4210138 | 1.0000000 | 0.00000 | 0.0104167 | 0.0251642 | 4T1_103 | H-2Kd | YYFTMAQQP |

Mupexi files

Load all mupexi files

mupexi_ct26 <- read_xlsx(path = "test_data/ct26_library_mupexi.xlsx") %>% 
  # remove extra columns from previous handling
  select(-identifier, -Mut_peptide.y, -Allele) %>% 
  # convert Mut_MHCrank_EL and Expression level to numeric so we can join both files
  mutate(Mut_MHCrank_EL = as.numeric(Mut_MHCrank_EL),
         Expression_Level = as.numeric(Expression_Level))

mupexi_4t1 <- read_xlsx(path = "test_data/4T1_library_mupexi.xlsx") %>% 
  select(-identifier)
kable(head(mupexi_4t1))

| HLA_allele | Norm_peptide | Norm_MHCrank_EL | Norm_MHCscore_EL | Norm_MHCaffinity | Norm_MHCrank_BA | Norm_MHCscore_BA | Mut_peptide | Mut_MHCrank_EL | Mut_MHCscore_EL | Mut_MHCaffinity | Mut_MHCrank_BA | Mut_MHCscore_BA | Gene_ID | Transcript_ID | Amino_Acid_Change | Allele_Frequency | Mismatches | peptide_position | Chr | Genomic_Position | Protein_position | Mutation_Consequence | Gene_Symbol | Cancer_Driver_Gene | Proteome_Peptide_Match | Expression_Level | Mutant_affinity_score | Normal_affinity_score | Expression_score | priority_Score | Self_Similarity | | :---------- | :------------ | :---------------- | :----------------- | :---------------- | :---------------- | :----------------- | :----------- | ---------------: | :---------------- | :--------------- | :--------------- | :---------------- | :-------------------- | :------------------------------------------ | :------------------ | :---------------- | :--------- | :---------------- | :-- | :---------------- | :---------------- | :-------------------- | :----------- | :------------------- | :----------------------- | ----------------: | :---------------------- | :---------------------- | :----------------- | :-------------- | :--------------- | | H2-Dd | SPPRFFHM | 0.0082 | 0.624879 | 491.4 | 0.0452 | 0.427228 | SPPRFFYM | 0.0084 | 0.616375 | 143.6 | 0.0132 | 0.540915 | MGP_BALBcJ_G0024183 | MGP_BALBcJ_T0049114,MGP_BALBcJ_T0049115 | H/Y | 0.525 | 1 | 7 | 18 | 29046509 | 13,13 | M | Wdr33 | - | No | 22.3967509 | 0.9999526549055792 | 0.9999527022247707 | 1.0 | 26 | 0.963257 | | H2-Kb | IAYAYTGNL | 0.0077 | 0.973579 | 2.8 | 0.0072 | 0.903204 | IAYAYRGNL | 0.0091 | 0.968819 | 2.9 | 0.0073 | 0.901354 | MGP_BALBcJ_G0030762 | MGP_BALBcJ_T0077749,MGP_BALBcJ_T0077750 | T/R | 0.179 | 1 | 6 | 6 | 54019866 | 91,91 | M | Kbtbd2 | - | No | 25.1951836 | 0.9999524889153081 | 0.9999528203159544 | 1.0 | 9 | 0.951350 | | H2-Db | TAIINHPLL | 0.0053 | 0.959733 | 21.7 | 0.0098 | 0.7156 | TGIINHPLL | 0.0111 | 0.911029 | 132.9 | 0.0664 | 0.548069 | MGP_BALBcJ_G0025205 | MGP_BALBcJ_T0052051 | A/G | 0.500 | 1 | 2 | 19 | 45414840 | 15 | M | Cfap43 | - | No | 44.9245980 | 0.9999520114438829 | 0.9999533830625381 | 1.0 | 25 | 0.974213 | | H2-Kd | SYTAPNTIF | 0.1807 | 0.502292 | 1298.5 | 1.0411 | 0.337423 | SYTAPNTIL | 0.0129 | 0.871451 | 45.5 | 0.0598 | 0.647101 | MGP_BALBcJ_G0026922 | MGP_BALBcJ_T0059497,MGP_BALBcJ_T0059498 | F/L | 0.346 | 1 | 9 | 2 | 171540148 | 269,215 | M | Gcnt7 | - | No | 0.1187235 | 0.9999515776185053 | 0.9998879552318652 | 0.215301102252408 | 4 | 0.985810 | | H2-Kb | HMFLFARL | 0.0064 | 0.978047 | 3.4 | 0.0085 | 0.885905 | HMFLFGRL | 0.0203 | 0.94733 | 5.9 | 0.0133 | 0.836579 | MGP_BALBcJ_G0026916 | MGP_BALBcJ_T0059483 | A/G | 0.467 | 1 | 6 | 2 | 171330471 | 214 | M | Mc3r | - | No | 0.1588522 | 0.9999497525243818 | 0.9999531259750578 | 0.2532216695531489 | 6 | 0.965400 | | H2-Dd | AGFSSFQKI | 0.0881 | 0.371051 | 12894.2 | 2.6668 | 0.125256 | AGFSSFQKL | 0.0267 | 0.465965 | 10559.4 | 1.8413 | 0.143719 | MGP_BALBcJ_G0031040 | MGP_BALBcJ_T0079002 | I/L | 0.483 | 1 | 9 | 6 | 90823751 | 1826 | M | Adamts9 | - | No | 2.3905895 | 0.9999481186865968 | 0.9999294769199811 | 0.9863617148770685 | 24 | 0.992681 |

Sample information file

sample_info <- read_xlsx(path = "test_data/sample_info.xlsx")
kable(head(sample_info))

| sample | percent_PE | | :------- | ----------: | | CT26_C1 | 4.23 | | CT26_C3 | 3.41 | | CT26_C4 | 2.79 | | CT26_D1 | 3.64 | | CT26_D2 | 3.22 | | CT26_D4 | 3.84 |

Merge all baracoda files

all_barracoda <- full_join(all_ct26_barracoda_raw, all_4t1_barracoda_raw)

Merge mupexi files

all_mupexi <- full_join(mupexi_4t1, mupexi_ct26) 

Merge data function

Use merge function to merge baracoda and mupexi files the output is my_data

my_data <- merge_all_data(barracoda =  all_barracoda,
                          mupexi  = all_mupexi,
                          info = sample_info)

Clean data function

With clean data function the varribale is cleaned and ready to import in augment_data function

my_clean_data <- clean_data(my_data)

Augment data

This Augment function will make new column with esimated frequency from the barracoda screening

my_clean_augment_data <- augment_data(my_clean_data)

Plotting functions

The packages contaions folowing plotting functions

barc_resp: Ilustrateing responses from baracoda

Firstly the responses can be explored in the following function, where the cell line and data is selected

 barc_resp(data = my_clean_augment_data,
           mouse_cell_line = "CT26") +
  labs(title = "log fold change of CT26-tumor cell line neoepitope screen")

scatterplot_function: Making scatter plot from data

To look closer to responses and discover the distribution of Imporved Binder (IB) and Conserved Binder (CB), the scatter plot funtion can be used.

scatterplot_function(data = my_clean_augment_data,
                           x = 'mut_mhcrank_ba', 
                           y= 'norm_mhcrank_ba')+
  labs(title= "Binding affinity (BA) of neoepitope vs WT epitope",
       x= "Neoepitope BA %Rank ",
       y="WT epitope BA %Rank")

bar_plot_func: Making bar plot from data

To investegate the muations possition in missense mutations the bar plot function is constructed

bar_plot_func(data = my_clean_augment_data,
                    pep_length = 10)

box_function: Making boxplot

To see wheter the is a diffrence in the responses and non responses to diffrent varraible the box plot function can be used

box_function(data = my_clean_augment_data, 
                   x = 'response',
                   y= 'self_similarity') +
  labs(title = "Self-similarity ", 
       x = "Response", 
       y = "Self Similarity",
       color = "Response")

Explore data in shiny app

Tha data can easy be open in a shiny app to explore the responses and the if there is and pattern in the immunugenic neoepitopes in Shiny_exploring Or with the following function in Rstudio.

Exploring_data_shiny(Plotting_data = my_clean_augment_data )

Shiny applications not supported in static R Markdown documents

rforbiodatascience/barcc documentation built on May 17, 2020, 5:31 p.m.