RADS (Rapid Alignment of Domain Strings) searches with a query in a database for similar domain arrangements. It has the following features:
This website covers the basics of RADS. For more detailed information please check the manual.
Although we try to keep the dependencies to a minimum some existing libraries are needed.
In most Linux distributions (e.g. Ubuntu, Arch Linux) it should be possible to install most of these dependencies using the package manager.
There are two ways to download RADS. Either you can download it using git or manually from the website. Both ways are described below.
If you use git you can easily update to a newer version when available.
Download using git
Use git to clone the repository and download BioSeqDataLib as a submodule:
git clone https://ebbgit.uni-muenster.de/domainWorld/RADS.git cd RADS git submodule init git submodule update
Inside the source folder a build directory is needed in which the code will be compiled. CMake is used to find all the needed requirements of the library.
mkdir build cd build cmake .. make
Sometimes it will be necessary to update RADS either because it contains some new features or because we unfortunately had a bug somewhere that we have now fixed. If you used git for the original download you can simply use git to update your code. Simply change into the RADS directory and type:
git pull git submodule foreach git pull origin master
You now simply follow again the steps in the Installation section. If you downloaded the code without git you will have to download the latest version and replace the old one with it, Do not forget to update the BioSeqDataLib folder as well.
You will need to have a Domain Similarity Matrix (DSM) installed in your system.
There are two possibilities to get a RADS database. The easiest is to use one of our precomputed matrices. If they do not contain the sequences you need you can very simply create your own database.
We provide some precomputed databases:
Currently we provide precomputed databases based on the InterPro domain annotations. With RADS version 2.3 we updated the database format. The old databases are not compatible anymore with the new version of RADS.
|interPro69-pfam.tar.bz2||864 MB (3.1G GB)||pfam-31.dsm||Contains all Pfam matches of the InterPro annotation (version 69).|
|interPro69-ssf.tar.bz2||777 MB (2.8G GB)||ssf-1_75.dsm||Contains all SuperFamily matches of the InterPro annotation (version 69).|
Creating your own database
It is very simple to create your own database. You need domain files in a supported format (e.g. the output of pfam_scan.pl). If you want to have sequence lengths given in the RADS output you will need to provide the sequences in fasta format as well. If you do not provide sequences, the length will be set to 0.
makeRadsDB -i domainFile1.pfam domainFile2.pfam -s seqFile1.fa seqFile2.fa -o myDB
The command above will create two files myDB.db and myDB.da. Both are needed by RADS.
After you have set up your system as described above you can get a short overview on how to use RADS to find similar domain arrangements. A more detailed description can be found in the manual.
You can provide the query in different formats:
rads --db interPro64-pfam -M pfam-31.dsm -D PF02758 PF05729
rads --db interPro64-pfam -M pfam-31.dsm -Q seq.fasta
rads --db InterPro60-pfam -M pfam-31.dsm -q seq.dom
The output of RADS consist of a single file containing the targets found by RADS (find below an example).
# RADS version 2.1.1 # RADS Output v1 # run at Thu Aug 3 15:59:19 2017 # # query file: - # database: interPro64-pfam # matrix: pfam-31.dsm # ****************************************************************** Results for: manual entered query Domain arrangement: PF00001 # score | normalized | SeqID | sequence length | domain arrangement # ------------------------------------------------------------------- 100 1.00 10020:000030 611 PF00001 44 293 100 1.00 10020:000054 276 PF00001 2 215 100 1.00 10020:0001c3 337 PF00001 42 293 100 1.00 10020:000327 402 PF00001 75 353 100 1.00 10020:000359 410 PF00001 52 305 100 1.00 10020:000393 372 PF00001 67 321
The targets are listed in a table consisting of five columns. Each column is tab separated from the next. The columns are:
If you find a problem, have questions or any kind of comment please contact us (domainworld[@]uni-muenster.de).
If you use RADS in your project please cite our publication:
Terrapon, Nicolas, Weiner, January, Grath, Sonja, Moore, Andrew D, Bornberg-Bauer, Erich: Rapid similarity search of proteins using alignments of domain arrangements., Bioinformatics (2014) 30 (2): 274-281. doi: 10.1093/bioinformatics/btt379