Creating a New Project: Delimited Text Files
A delimited text file is simply any plain text file that contains a "table" of data in which each column is separated by a certain character. This type of file is extremely common; most spreadsheet applications can create them, and R and other statistical analysis packages can output results to this type of file. For this example, we'll be looking at the following file, created in Excel:
To save an Excel file as a delimited text file, choose File->Save As... In the dialog that appears, choose either the Text (Tab delimited) or CSV (Comma delimited) option from the Save as type drop-down menu. Note that whatever character you use as a delimited can't appear anywhere in the data (i.e. don't choose Comma delimited if any commas appear in your data).
Note that data can come from any source, not just Excel. In fact, for extremely large data sets (which MetaOmGraph is adept at handling), Excel may be unusable. An Excel file is simply used as an example. The process for data from any other source is identical.
Loading Your Data
Before loading your data into MetaOmGraph, there are three things you need to know about your file:
-
Delimiter: the delimiter character used to separate columns in your file. For example, if you selected
Text (Tab delimited), the delimiter is the Tab character. Similarly, if you're working with a CSV file, your delimited will be the comma. - Blank values: whether or not there are any blank values/empty cells in your data. Unless the value of a certain cell is truly "blank" (i.e. no value exists, rather than the value is simply zero or below a detection value), it is usually best to fill every blank cell with a more meaningful value (such as zero). MetaOmGraph can handle blank values, but they will slow down correlations, and results may be less meaningful.
- Info columns: the number of columns at the start of each row of data used to identify that row. In the example above, there are three info columns: ExperimentID, RowID, and MetaboliteName. Data in info columns will only be used to identify rows. It won't appear in created plots, nor will it be a factor in correlations.
When you're ready to load your file into MetaOmGraph, choose File->New Project->From Delimited Text File... In the New Project dialog, click the Browse... button next to Data File and select your data file. MetaOmGraph will analyze the file, and if successful, will fill in the rest of the fields for you. The status message at the bottom of the dialog will indicate whether the analysis was successful. Check the rest of the fields to ensure they're correct, then click OK. Your data will be loaded into MetaOmGraph.
Retrieving Data From MetNetDB
If your info columns/row identifiers contain arabidopsis gene IDs (locus IDs, affy IDs), after loading your project, you will be prompted to connect to MetNetDB to retrieve additional data. This data includes gene names, and MetNet pathways in which the genes appear. Retrieving this data is optional, and requires an Internet connection.
Optional Files
There are three optional files that you can load in addition to your data file.
- Row Names file: if your row identifiers/info columns don't appear in the original data file, and instead are in a separate text file, you can load that file here. The file should consist of a header followed by row identifiers for each row, each appearing a separate line.
- Column Names file: if you want to replace the header row in your data file with a different set of headers, you can load those here. The file should contain a new header for each column, one per line.
- Metadata file: if you have metadata for your data file, you can load it here. See the metadata page for more information.