[ RADS ][ RADS ]

Please note: This webapplication is a proof-of-concept of the RADS-RAMPAGE algorithms (publication in Bioinfomatics in January 2014). When using it, please keep in mind that it is still under construction.

Batch searches: For running RADS/RAMPAGE searches in batch mode, we provide a command-line program which queries the web interface. It can be downloaded here (requires Java).

User id 1467444074-6337

[ Move your mouse over the little yellow buttons to get context help ]


RADS Search

Search for homology using domain-wise arrangements

[ Reset form ] [ Control panel ]

Query (identifier, sequence or XDOM):

Database:

Domain - dot plots

Visually compare two proteins using a dotplot showing annotated domains

[ Reset form ]

Database:

First query (identifier, sequence or XDOM):

Second query (identifier, sequence or XDOM):





















Documentation

What is RADS?

For a quick start, use the green [ EXAMPLE ] buttons.

RADS is a program to detect homology of proteins by comparing the domain arrangements of proteins. In RADS, amino acid sequences are not directly compared between your query and the proteins in the database
(here, Simap that gather proteins from PDB, Uniprot, RefSeq and Genbank). Instead, RADS examines the protein domains (from Pfam database) present in the protein and looks up in our local database which proteins
have a similar arrangement of domains.

For help, move your mouse over the yellow question marks: [ ? ]

This approach has two major advantages.

  • The detection of protein domains is more sensitive than comparison between two sequences (pairwise), because it relies on comparing a sequence to a set of sequences (family model/profile). Hence, RADS
    allows to detect distant homologies that you would normally be able to find only by using profile BLAST (PSI-BLAST) or similar methods.
  • Once the domains in your query are determined, the homology search with RADS is much faster than a comparison of amino acid characters (shortened sequences).

What is RAMPAGE?

RAMPAGE uses RADS to create a pairwise alignement. First, matching domains are found by RADS. Then, RAMPAGE aligns first the matched domains, and then all other regions of the two sequences. Finally, the aligned fragments are put together.

Finally, using RADS along with BLAST, allows you to find functional homologs that have an identical (or similar) domain arrangement, but little sequence conservation. To identify such case, you could look for the
hits with low BLAST score (or those which were not detected by BLAST), but having a high RADS score.


Running RADS

RADS needs only information about domains, RAMPAGE also needs information about a sequence. Both are happy with just an identifier from the database used.