DAPPLE represents an alternative method (to machine-learning approaches) to predicting phosphorylation sites in an organism of interest. It is a pipeline involving BLAST searches that uses experimentally-determined phosphorylation sites in one organism (or several organisms) to predict phosphorylation sites in an organism of interest. It outputs a table in tab-deliminated text format (which can also be easily imported into a spreadsheet program like Excel), which contains various information helpful for choosing phosphorylation sites that are of interest to you, such as the number of sequence differences between the query site and the hit site, the location of the query site and the hit site in their respective intact proteins, whether the corresponding intact proteins are reciprocal BLAST hits (and thus predicted orthologues), and so on.
The following is a web interface to DAPPLE. If you would instead like to run DAPPLE on your own machine, you may download it here. This .zip file includes instructions for setting up DAPPLE.