December 20, 2005

What is PathBlast?

PathBlast is general strategy for aligning two protein interaction networks to elucidate their conserved pathways. This method identifies pairs of interaction paths, drawn from the networks of different species or from different processes within a species, where proteins at equivalent path positions share strong sequence homology.

What can I search against?

The current version of PathBLAST searches against known protein interactions collected in the October 23, 2005 version of the DIP protein interaction network database.

With what can I query?

The web-version of PathBlast can be queried in two ways, either by ID or by sequence. Valid ID's consist of any synonym found in the DIP interaction database. If an ID is not an exact match, we'll present you with potential matches in the database that you can choose between. If you choose to search by sequence, then any valid protein sequence is acceptable. If both ID and sequence are entered the ID will be ignored and used only as an identifier. The actual search will take place using the sequence. The query path is inferred from the order that the sequences are entered. Each query can consist of two to five protein sequences.

How is PathBLAST implemented?

With the exception of the BLAST code, PathBLAST is implemented entirely in Java. Details of the various algorithms and techniques used can be found in the various publications that describe different aspects of PathBLAST.

Can I download PathBLAST?

Hopefully soon....

Help, PathBLAST returned an error!

Obviously, this shouldn’t happen. Please email a copy of the error message to and we’ll fix it as soon as possible.

I want to submit a larger query.

For performance reasons, the web version is limited in the size of the query graphs. If you would like to run larger networks, please see the section above about downloading PathBLAST

Why do my queries look different from the PNAS paper?

PathBLAST has been completely re-implemented and now uses Cytoscape to render the graphs. In addition various algorithmic improvements have been made. See the publications section for more details.

Can I compare the results of a query against Yeast to a query against Fly?

Of course. The one caveat is that you can only do this for a specific version of PathBLAST. E-value scores may change when the size of the Blast protein database changes. Updates to PathBLAST change the size of this protein database and hence change the E-values returned. If you want to compare queries against multiple species, simply make sure that all queries all performed using the same version of PathBLAST.