RADIX is software for identifying exons which show differential rates of inclusion within a gene between different sample types. It's core model is built around a beta-binomial regression, allowing it to support multiple replicates per sample type, and arbitrary experimental design.
Copyright (C) 2014 University of Southern California, Philip J. Uren, Andrew D. Smith, Egor Dolzhenko
Authors: Philip J. Uren, Andrew D. Smith, Egor Dolzhenko
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.
This software package contains Google Test and BAMTools -- see their respective directories for copyright and license information for those packages.
RADIX has been designed to operate in a UNIX-like environment. It has been tested on MacOS X Mavericks, USCLinux, Ubuntu and SUSE. We've made every effort to ensure it's compatible with a wide range of systems, but we can't test on all possible platforms and configurations. If you find an incompatibility when building or using it, please let us know!
64-bit machine and GCC version > 4.1 (to support TR1).
RADIX additionally requires a functioning installation of the GNU Scientific Library (GSL). If you don't already have this installed, you will need to download and install it from http://www.gnu.org/software/gsl/ This release has been tested with GSL version 1.14 Your installation of GSL must be built for a 64 bit architecture.
[OPTIONAL] Tests The regression tests (which you don't need to run, but can to test that RADIX was built correctly) require Python. If you don't have it, you can download Python from http://www.python.org/ This release has been tested with Python 2.6.1
First configure the installation. To do this, change into the root directory of the distribution and type:
./configure
To build the binaries, again in the root directory of the distributon, type:
make all
To install the binaries, again in the root directory of the distribution type:
make install
This will place the binaries in the bin directory under the package root. They can be used directly from there without any additional steps. You can add that directory to your PATH environment variable to avoid having to specify their full paths, or you can copy the binaries to another directory of your choice in your PATH.
You can verify that RADIX was built correctly by running the included unit tests and regression tests. In the root directory of the distribution type:
make test
RADIX takes as input a set of read counts for exons in each sample
and a design matrix describing the experimental setup. These two inputs
are always required, and in the simplest use of the program nothing else
need be specified. For example, if your design matrix was in the file
design.mat and you had four replicates (say two for some experimental
condition and two control replicates) with read counts in
control_rep_1.bed, control_rep_2.bed, exp_rep_1.bed and
exp_rep_2.bed then you would run the program as follows
radix -d design.mat control_rep_1.bed control_rep_2.bed exp_rep_1.bed exp_rep_2.bed
Note that if you have more than one condition, you must also specify the option '-f ' to indicate which condition you wish to test. See below for more details about conditions and how to specify them in the design matrix. Additional options can be specified to adjust, for example the signficance threshold used by RADIX. Run the program with no arguments to see a list of other options that can be used:
radix
Readcounts for exons are provided to RADIX in BED format files. The counts for each sample are provided in a seperate file. These files must contain the following tab-seperated columns: chromosome, exon start, exon end, exon name, score (the count of reads in this exon), and exon strand (+ or -). Here are some example entries:
chr1 1334022 1334345 gene1_exon_1 34 +
chr1 1334808 1334899 gene1_exon_2 10 +
chr3 124904 1250023 gene2_exon_1 24 -
exons must follow this naming style: geneName_exon_exonName, where exonName and geneName can be any string desired, but must form a unique identifier. All entries with the same value for 'geneName' are considered to be part of the same gene. The exons for a given gene must be non-overlapping, and all input files must contain read-counts for the same set of exons. If either of those conditions is not met, the program will exit with an error. BED files are tab seperated.
The first row of the design matrix must contain each of the experimental conditions in your data, expressed in such a way that you can answer the question of whether that condition is present or not in each sample. As a simple example, consider testing two control replicates against two cancer replicates. Here, there is just one condition: cancer, and each sample either has, or does not have this condition (note that there is no need to include "control" as a condition since it is just the inverse of the "cancer" condition). Each subsequent line in the file should start with a filename that contains the exon read counts for a sample and be followed by either a zero or a one indicating whether the corresponding condition is present for that sample or not (i.e. the first column after the name corresponds to the first condition listed in the file's first line, the second column corresponds to the second condition listed in the file's first line, etc.) Here's an example:
some_experimental_condition
control_rep_1.bed 0
control_rep_2.bed 0
exp_rep_1.bed 1
exp_rep_2.bed 1
Exact alignment and additional whitespace does not need to be maintained, we use it here just for display purposes.
RADIX produces output in BED format. A single BED file is produced, regardless of how many samples are provided. Each exon present in the input files will also be present in the output. The score field of the BED file (the fifth column) will be the mean log fold-change for the exon's inclusion rate. There will be an addition field (beyond the default six columns in a BED file) that gives the p-value the resulted from testing whether the exon was differentially included or not. Only exons with a p-value less than the significance threshold (set by default to 0.01, but adjustable) are output.
Philip J. Uren uren@usc.edu
Egor Dolzhenko dolzhenk@usc.edu
Andrew D. Smith andrewds@usc.edu
If you found a bug in RADIX, we'd like to know about it. Before contacting us though, please check the following list:
- Are you using the latest version? The bug you found may already have been fixed.
- Check that your input is in the correct format and you have selected the correct options.
- Please reduce your input to the smallest possible size that still produces the bug; we will need your input data to reproduce the problem.