Identification of Pak kinase phosphorylation sites


This method is based on our studies published here:

Rennefahrt et al, J. Biol. Chem. 282(21): 15667-15678.

We have characterized the kinase specificity of multiple Pak isoforms including Pak1, Pak2, and Pak4 using a degenerate peptide-based methodology described by Dr. Benjamin Turk in Nature Protocols 2006.

The resulting comprehensive kinase specificity matrix can be used to analyze each serine or threonine in a putative substrate to quantitatively assess how well each serine/threonine in its particular sequence context matches the preferred kinase consensus. The consensus matrices for the Pak isoforms can be downloaded as Excel spreadsheets below. Note that threonine and serine residues must be evaluated separately and scores for serine and threonine are not comparable. However, Pak generally exhibits a 5-10-fold preferred for serine over threonine as phosphoacceptor (Rennefahrt et al. 2007).


Download Pak specificity matrices

Pak1 threonine matrix
Pak1 serine matrix
Pak2 threonine matrix
Pak2 serine matrix
Pak4 threonine matrix
Pak4 serine matrix

How to use the specificity matrices to identify Pak phosphorylation sites:

1) Click on the link above and save the desired Pak specificity matrices to your computer.
2) Navigate to
3) Under "motif scan" select "Scan Input Sequence with an Input Motif"

4) Enter a name for the motif (e.g. Pak1 threonine matrix) to be used .
5) Click "choose file" and upload the Pak specificity matrix. Click "submit"

6) You will be presented with a view of uploaded file for verification. Click on "Yes, I would like to continue with this matrix"
7) Enter a name for the protein sequence to be scanned and paste the raw amino acid sequence into the text box. Click "submit request"

8) A graphical representation of the protein is shown in which all serines or threonines (depending on the matrix used) are numberd. A second graphic shows plot of predicted surface accessibility along the length of the query protein sequence.

Below, all serines (or threonines) are listed from those BEST matching the Pak consensus (at the top) to WORST (the bottom). A lower scansite "score" indicates a BETTER match to the Pak consensus. The sequence surrounding each site is given along with the predicted surface accessibility for that site (SA).

As a guideline, sites with Scansite scores of <1 should be considered as candidate Pak sites although some bona fide Pak sites can have higher scores. For more guidance in choosing candidate Pak substrate sites see Rennefahrt et al. 2007. Details about the excellent Scansite tool can be found here.