In espors/idepGolem: Integrated Differential Expression and Pathway Analysis

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
options(rmarkdown.html_vignette.check_title = FALSE)

# library(idepGolem)
devtools::load_all()

This instruction will document how to Load Data and create the data objects that will be used in the rest of iDEP's functions. This corresponds to the functions that are utilized in the Load Data tab of the iDEP webpage. The step-by-step will guide you through loading and running the initial functions to set up the rest of iDEP analysis.

Initial Data Paths

To begin working with iDEP the first step is to identify your working directory and set it to the path where your data is located. The first two files that will be worked with is the expression data file and the file containing information on the experiment design and groupings for the samples. The design for the expression data file should have the gene IDs as the first column. The ensuing columns should each correspond to a sample, with the name of the sample in the first row. The data should then be then correspond (row, column) as the count for (gene, sample). If an experiment data file is included for the data, the structure should be similar to the expression data file. The first column should be the name of an experiment effect. The ensuing column then label each specific sample. The data will then correspond (row, column) as the group for a (effect, sample). This instruction will use the built-in demo expression and experiment files. The first code block will set the data paths to these files which will be filled into the first functions. We will also run the get_idep_data function to return a list of the data objects used in the iDEP data analysis.

idep_data <- get_idep_data()

DATABASE <- Sys.getenv("GE_DATABASE")[1]
YOUR_DATA_PATH <- paste0(DATABASE, "data_go/BcellGSE71176_p53.csv")
YOUR_EXPERIMENT_PATH <- paste0(DATABASE, "data_go/BcellGSE71176_p53_sampleInfo.csv")


expression_file <- data.frame(
  datapath = YOUR_DATA_PATH
)
experiment_file <- data.frame(
  datapath = YOUR_EXPERIMENT_PATH
)

input_data Function

This function will read the data from the given data paths and return a list with the data objects. data will be the data matrix of expression data that was detailed in the paragraph above. data will have row names as the gene IDs and column names matching the names of the samples that each column represents. sample_info will be the experiment design information that was detailed above. The structure will be the transpose of the original file structure, with the the rownames now being the column names of the original file. The column names of this data matrix will label an effect in the experiment. For example, in the demo data sample_info, the column names are ("p53", "Treatment") for the muted gene and radiation treatment. If there is no experiment file, the function will only return the expression data.

For this example, we will store the data in a list called load_data.

load_data <- input_data(
  expression_file = expression_file,
  experiment_file = experiment_file,
  go_button = FALSE,
  demo_data_file = idep_data$demo_data_file,
  demo_metadata_file = idep_data$demo_metadata_file
)

iDEP Database Functions

The next step of process is to use the gene IDs from the expression data to get the gene information and determine the species that the expression samples are from using the iDEP database. The following functions will also convert the IDs, if they are recognized, to ensembl and symbol name. These functions have an input parameter to specify the species that the data is for. The line below will shows how to search iDEP to find the numeric code for the species you wish to use. To allow iDEP to find the species, use "BestMatch". To use a numeric code for a specfic species, make sure to use string quotes, for example "107".

# Species search for input data
idep_data$species_choice[grep("Mouse", names(idep_data$species_choice))]

Each iDEP database function is outlined below.

1. convert_id

   This function will take in a 'query' and use the 'idep_data' to match the provided gene IDs with the corresponding ensembl ID. The IDs from the inputted data are currently the rownames of the data frame data stored in the list load_data. select_org is a parameter to provide the species that the expression data is from. convert_id returns a list containing information about the query results. The converted IDs are are accessed by converted$ids. print(converted$species_matched[1,]) shows the species that was recognized by the provided gene IDs.

converted <- convert_id(
  query = rownames(load_data$data),
  idep_data = idep_data,
  select_org = "BestMatch"
)

print(converted$species_matched[1, ])

1. gene_info

   Gathering the information on the converted ensembl IDs is the next step. The converted results are passed into the gene_info function to retrieve gene name information that is stored in idep_data as well. If select_org is "BestMatch", the function will use the matched species from converted to retrieve the gene data. If a species code is provided, it will use that species directly. gene_info will then read a gene data file from the idep_data and return the resulting data frame. This data will be used to map to more informative gene names in the coming functions.

all_gene_info <- gene_info(
  converted = converted,
  select_org = "BestMatch",
  idep_data = idep_data
)

1. convert_data

   This step will return the original data from input_data with the converted ensembl IDs as rownames, and a data frame of the mapped IDs. It uses the IDs that were matched with convert_id to swap the current rownames with the corresponding ensembl ID. no_id_conversion can be set to TRUE if the desired IDs are the original ones, but the simpler solution is to simply not run the function. The returned list contains the data with converted IDs in data and IDs in mapped_ids.

converted_data <- convert_data(
  converted = converted,
  no_id_conversion = FALSE,
  data = load_data$data
)

1. get_all_gene_names

   This function combines all the gene name information that has been gathered in the preceding functions. mapped_ids are the matched IDs from converted, and all_gene_info is the data read from idep based off the identified species. This function returns a data frame with User_ID as the original IDs, ensembl_ID containing the converted IDs, and symbol containing the informative gene names. This data frame can be used throughout the package to matach gene IDs.

gene_names <- get_all_gene_names(
  mapped_ids = converted_data$mapped_ids,
  all_gene_info = all_gene_info
)

Conclusion

This concludes the function calls and processes from the first tab of the iDEP program. There is no real data manipulation that happens, but the objects that are created are essential for the processes that will happen in the following instructions. For troubleshooting, all functions have documentation and the code is available on Github. (https://github.com/gexijin/idepGolem)

espors/idepGolem documentation built on Oct. 27, 2024, 4:56 a.m.