Grand Forest • Expression data file format

Overview

Gene expression data can be uploaded to the Grand Forest online platform in form of a data table. Delimited plain-text files are supported as well as R data objects. The file extension of the uploaded file is used to determine how to parse read the file so make sure to file is named accordingly.

The expression data file must fulfill the following requirements:

• The file must contain a valid header, mapping columns to genes in the network, and naming clinical variables such that they can be referred to.
• Each row corresponds to a sample/patient.
• Gene identifiers must be Entrez gene IDs. If your data uses different identifiers, you will need to map them to Entrez IDs first.
• All column names must be unique.
• All expression values must be numeric.

1. Comma-separated values (.csv)

Data can be uploaded as comma-separated values. Example of a valid CSV file:

label,2099,351,7534,8452
case,0.40,0.50,0.88,0.81
control,0.42,0.95,0.31,0.15
control,0.91,0.99,0.88,0.57
case,0.72,0.19,0.32,0.16
case,0.19,0.85,0.03,0.37
control,0.97,0.78,0.96,0.18

Values containing commas must be enclosed in quotes:

group,2099,351,7534,8452
"case,day1",0.40,0.50,0.88,0.81
"case,day1",0.42,0.95,0.31,0.15
"case,day7",0.91,0.99,0.88,0.57
"case,day7",0.72,0.19,0.32,0.16
"control,day1",0.19,0.85,0.03,0.37
"control,day1",0.97,0.78,0.96,0.18

2. Tab-separated values (.tsv)

Data can be uploaded as tab separated values. Example of a valid TSV file:

label	2099	351	7534	8452
case	0.40	0.50	0.88	0.81
control	0.42	0.95	0.31	0.15
control	0.91	0.99	0.88	0.57
case	0.72	0.19	0.32	0.16
case	0.19	0.85	0.03	0.37
control	0.97	0.78	0.96	0.18

Values containing tabs must be enclosed in quotes:

group	2099	351	7534	8452
"case	day1"	0.40	0.50	0.88	0.81
"case	day1"	0.42	0.95	0.31	0.15
"case	day7"	0.91	0.99	0.88	0.57
"case	day7"	0.72	0.19	0.32	0.16
"control	day1"	0.19	0.85	0.03	0.37
"control	day1"	0.97	0.78	0.96	0.18

3. Compressed files (.zip)

You can reduce the time it takes to upload your data by compressing the data file as a ZIP archive.

• In order for ZIP files to be recognized, the name of the uploaded file must have the .zip extension.
• If the archive contains more than one file, Grand Forest will read only the largest one.

4. R data single object files (.rds)

Data tables can also be uploaded as R data single object (RDS) files exported from the R programming language. For more information on RDS files see the R documentation. In order for Grand Forest to recognize the RDS format, the uploaded file must have .rds as file ending (e.g. mydata.rds).

The data frame must have valid column names set:

> print(X)
         group 2099  351 7534 8452
1    case,day1 0.40 0.50 0.88 0.81
2    case,day1 0.42 0.95 0.31 0.15
3    case,day7 0.91 0.99 0.88 0.57
4    case,day7 0.72 0.19 0.32 0.16
5 control,day1 0.19 0.85 0.03 0.37
6 control,day1 0.97 0.78 0.96 0.18

If your data frame is stored in a variable X, you can save it as an RDS file using the saveRDS-function:

saveRDS(X, file="data.rds")

The saved object must be of class data.frame or data.table. You can verify this with the class-function:

> X <- data.frame(...)
> class(X)
[1] "data.frame"

> Y <- data.table(...)
> class(Y)
[1] "data.table" "data.frame"