Start by uploading your statistics table



Download example | Run example

PinPath v0.1.0 | MIT license


Save

Transparency (0-1)
Node size (0-5)
Network layout
Show unconnected nodes



About PinPath

PinPath can be used to visualize data onto pathway diagrams, and pinpoint where in the pathway the relevant changes occur. Results from (epi)genomics, transcriptomics, (phospho)proteomics, metabolomics and many more experiments can be visualized onto pathway diagrams from KEGG and WikiPathways. You can also upload your own GPML and KGML files to visualize data onto custom pathways. As long as your data can be linked to genes, proteins, or metabolites, you can visualize it using PinPath.

Table of content

  1. Data upload
  2. Pathway visualization
  3. Overrepresentation analysis
  4. Contact

Data upload

The uploaded data should include at least two columns: one column with the gene, protein, or metabolite IDs and one column that can be used for coloring. Including more columns is also possible, as long as there is at least one column with feature IDs and at least one column that can be used for coloring.

This is how a statistics table could look like:

Gene     log2FC       p-value
MECP2     2.24       0.0001
LHX1     1.23       0.0345
FUT5     0.98       0.6418
….     ….       ….
If your statistics table looks like this, you can later decide to color the pathway nodes with the log2FC, the p-value, or both!

Pathway visualization

After you have uploaded the data, you can visualize the data onto the pathway. Particularly, you can color the nodes of the pathway by one or more variables. The color palette can be fully customized and the pathway image, legend image, and node table can be downloaded as separate files. Furthermore, all pathways diagrams can also be converted to a network visualization.


Overrepresentation analysis

You can also perform overrepresentation analysis to identify relevant pathways. For this, you need to have column in the statistics table that indicates whether a gene is differentially expressed or not. If this column does not exist, you can add this column (see Adding columns to data tab). The clusterProfiler package is used for the overrepresentation analysis. Note that overrepresentation analysis on the metabolites is not supported (yet).


Contact

Adding columns to data

After you have uploaded your data, you can add columns to your data using the + button in the Data info page. You can add a column that discetizes or transforms an existing variable in the data.

Discretization

The rule system for discretizing a contiuous variable (e.g., p-value or logFCs) consist of the following structure:

Label 1: rule that defines label 1
Label 2: rule that defines label 2
...
Label n: rule that defines label n

For instance, you can add a column to the data that indicates whether a gene is significantly differentially expressed:

Yes: `p-value` < 0.05
No: `p-value` >= 0.05

Here is another example to make an extra column that indicates whether a gene is significantly up- or downregulated:

Up: (`Significant` == "Yes") & (`log2FC` > 0)
Down: (`Significant` == "Yes") & (`log2FC` < 0)
Unchanged: `Significant` == "No"

Note in these examples that the column name is put in-between backticks (e.g., `column name`).

Transformation

You can also transform a continuous variable. For instance, perform log10-transformation on the p-value:

log10(`p-value`)

Moreover, you can convert the log2FC to the fold change:

2^`log2FC`

Session information

Download