In tidymass/massdataset: massdataset

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warning = FALSE,
  message = TRUE,
  out.width = "100%"
)

Data preparation

The massdataset class object can be used to store the untargeted metabolomics data.

Let's first prepare the data objects according to the attached figure for each file.

1. sample_info (required)

The columns sample_id (sample ID), injection.order (injection order of samples), class (Blank, QC, Subject, etc), group (case, control, etc) are required.


2. variable_info (required)

The columns variable_id (variable ID), mz (mass to charge ratio), rt (retention time, unit is second) are required.


3. expression_data (required)

Columns are samples are rows are features (variables).

The column names of expression_data should be completely same with sample_id in sample_info, and the row names of expression_data should be completely same with variable_id in variable_info.


4. sample_info_note (optional)

This is the metadata for sample_info.


5. variable_info_note (optional)

This is the metadata for variable_info.


Download demo data

Here we use the demo data from masssprocesser package. The demo data can be downloaded here.

Download this data and uncompress it. And then set the path where you put the folder as working directory.

Then prepare data.

library(tidyverse)
peak_table_pos = readr::read_csv("feature_table/Peak_table_pos.csv")
peak_table_neg = readr::read_csv("feature_table/Peak_table_neg.csv")

sample_info_pos = readr::read_csv("feature_table/sample_info_pos.csv")
sample_info_neg = readr::read_csv("feature_table/sample_info_neg.csv")

Variable information and expression data are in the peak table. Let's separate them.

expression_data_pos = 
  peak_table_pos %>% 
  dplyr::select(-c(variable_id:rt)) %>% 
  as.data.frame()

variable_info_pos = 
  peak_table_pos %>% 
  dplyr::select(variable_id:rt) %>% 
  as.data.frame()

rownames(expression_data_pos) = variable_info_pos$variable_id

expression_data_neg = 
  peak_table_neg %>% 
  dplyr::select(-c(variable_id:rt)) %>% 
  as.data.frame()

variable_info_neg = 
  peak_table_neg %>% 
  dplyr::select(variable_id:rt) %>% 
  as.data.frame()

rownames(expression_data_neg) = variable_info_neg$variable_id

colnames(expression_data_pos) == sample_info_pos$sample_id
colnames(expression_data_neg) == sample_info_neg$sample_id

The orders of sample_id in sample_info and column names of expression_data are different.

expression_data_pos = 
  expression_data_pos[,sample_info_pos$sample_id]

expression_data_neg = 
  expression_data_neg[,sample_info_neg$sample_id]

colnames(expression_data_pos) == sample_info_pos$sample_id
colnames(expression_data_neg) == sample_info_neg$sample_id

Create mass_data class object

Then we can create mass_data class object using create_mass_dataset() function.

library(massdataset)

object_pos =
  create_mass_dataset(
    expression_data = expression_data_pos,
    sample_info = sample_info_pos,
    variable_info = variable_info_pos
  )

object_pos

Then negative mode.

object_neg =
  create_mass_dataset(
    expression_data = expression_data_neg,
    sample_info = sample_info_neg,
    variable_info = variable_info_neg
  )

object_neg

Then save them for next analysis.

save(object_pos, file = "feature_table/object_pos")
save(object_neg, file = "feature_table/object_neg")

Export mass_dataset class object to csv for xlsx

export_mass_dataset(object = object_pos,
                    file_type = "xlsx",
                    path = "demo_data_pos")

Then all the data will be in the demo_data_pos folder.


mzMine feature table to mass_dataset class

We can also directory convert feature table from mzMine to mass_dataset class.

An example feature table from mzMine.

data("mzmine_table")
library(kableExtra)
library(magrittr)
kbl(mzmine_table) %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"),
                fixed_thead = TRUE) %>% 
  scroll_box(width = "100%", height = "600px")

data("mzmine_table")
object =
   convet_mzmine2mass_dataset(x = mzmine_table)
object

MS-DIAL feature table to mass_dataset class

We can also directory convert feature table from MS-DIAL to mass_dataset class.

An example feature table from MS-DIAL.

data("msdial_table")
library(kableExtra)
library(magrittr)
kbl(msdial_table) %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"),
                fixed_thead = TRUE) %>% 
  scroll_box(width = "100%", height = "600px")

data("msdial_table")
object =
   convert_msdial2mass_dataset(x = msdial_table)
object

Session information

sessionInfo()

tidymass/massdataset documentation built on Jan. 30, 2024, 2:55 p.m.