Chapter 1 📥 Load data into R environment

1.1 Organization of built-in data

The built-in data follows several specifications for constructing:

① They are constructed into a SummarizedExperiment object using the SummarizedExperiment() function in SummarizedExperiment package.

② The assays of the SummarizedExperiment object are stored as a SimpleList containing an expression matrix, where rows represent genes and columns represent samples.

③ The colData of the SummarizedExperiment object is a table(DFrame object of S4), where the row names must be the same as the column names of the expression matrix. The column names of the table represent the clinical information of patients. The detailed information about the column names is presented below.

Column name Recommend value Description Necessity in tigeR analysis
sample id character The sample identity. Required
patient_name character The patient identity. Alternative
dataset_id character The dataset identity. Alternative
Treatment PRE/POST The patient sample is collected after Treatment or before Treatment Required
response CR/MR/PR/PD/SD/NR/R/N The immunotherapy response. Alternative
response_NR R/N The immunotherapy response which only contains R or N. Required
M.stage character The M stage of patients. Alternative
overall.survival..days. value The number of days, months or years of survival time of patients.(all samples must be the same unit) Required (for survival analysis)
vital.status Alive/Dead The survival status of patient. Required (for survival analysis)
Total.Mutation numeric The total mutation gene numbers. Alternative
Gender M/F The gender of patient. Alternative
Therapy character The anti-tumor therapy on patients. Alternative
age_start numeric The age of the patient at diagnosis. Alternative
tumor_type character The type of tumor. Alternative
seq_type character The sequencing type. Alternative
id character The identity of dataset. Alternative

1.3 Obtain data from tigeR web server

Dataloader(c(1,2,3), use_source="Web Server")

1.4 Obtain data from ExperimentHub

Dataloader(c(1,2,3), use_source="ExperimentHub")

1.5 Pre-processing of custom Data

 When conducting analysis using custom data, you need to pre-process your data and construct a SummarizedExperiment object. Prepare at least one gene expression matrix (rows for genes, columns for samples) and a data frame including corresponding clinical information of the samples. The data frame should include at least the following information:

  • sample id
  • Treatment (PRE/POST)
  • response_NR (R/N)
  • overall.survival..days. (for survival analysis)
  • vital.status (for survival analysis)

 Here is a brief example:

## Construt FPKM Matrix
genes <- c("5S_rRNA", "A1BG", "A1BG-AS1")
samples <- c("SRR3184279", "SRR3184280", "SRR3184281")

mtr <- matrix(c(0.04115157, 0.01311394, 0.02275499,
                0.07774626, 0.29494916, 0.11743900,
                2.01782650, 0.83763027, 0.77378394),
              nrow = 3, ncol = 3, byrow = TRUE, 
              dimnames = list(genes, samples))

## Construct Clinical Information Table
data <- data.frame(
  sample_id = c("SRR3184279", "SRR3184280", "SRR3184281"),
  Treatment = c("PRE", "PRE", "PRE"),
  response_NR = c("N", "R", "R"),
  overall_survival_days = c(607, 927, 948),
  vital_status = c("Dead", "Alive", "Alive"),
  stringsAsFactors = FALSE)

## Construct SummarizedExperiment Object
SE <- SummarizedExperiment::SummarizedExperiment(assays = list(fpkm = mtr),
                                                 colData = data)

SE
# class: SummarizedExperiment
# dim: 3 3
# metadata(0):
#   assays(1): fpkm
# rownames(3): 5S_rRNA A1BG A1BG-AS1
# rowData names(0):
#   colnames(3): SRR3184279 SRR3184280 SRR3184281
# colData names(5): sample_id Treatment response_NR overall_survival_days
# vital_status