SCHOOL OF BIOLOGICAL SCIENCES

NCBI

E-Services

test

SSP is a de novo transcriptome assembler that assembles RNA-seq reads into transcripts. SSP aims to reconstructs all the alternatively spliced isoforms and estimates the expression level of them.

· Installing a pre-compiled binary release

In order to make it easy to install SSP, we provide a few binary packages. To use the binary packages, simply download the appropriate one for your machine, untar it, and use it.

Linux x86_64 binary

Mac OS X x86_64 binary

Building SSP from source

SSP runs on Linux or Mac OS X based computers with GCC 4.0 or greater.

In order to execute SSP, you must follow below steps to have all the needed requirements installed on your system.

· Installing Integer Linear Programming solver

1. Download LP_solve

2. Unpack the lp_solve tarball

· Installing Velvet

1. Download Velvet

2. Unpack the Velvet tarball and cd to Velvet source directory

3. Build the Velvet by typing make at the command line.

· Installing Clustering_SSP

1. Download Clustering_SSP

2. Unpack the Clustering_SSP source tarball:

tar Clustering_SSP.tar.gz

3. Change to the Clustering_SSP directory:

cd Clustering_SSP_0.1

4. Make the Clustering_SSP :

make

· Building SSP

1. Download SSP

2. Unpack the SSP source tarball:

tar zxvf SSP_Source.tar.gz

3. Change to the SSP directory:

cd SSP_0.1.01

4. Configure SSP. You need to specify where to find lp_solve, Velvet and Clustering_SSP binary files:

./configure.sh -lp_solve=/path/to/lp_solve -Clustering_SSP=/path/to/ Clustering_SSP

Running SSP

Run SSP from the command line via the script SSP.sh, which is available in the base installation directory.

1- Preprocess the RNA-seq reads using Velvet.

a. It’s necessary to run velvetg by setting the value of -read_trkg parameter to yes. Do not set -cov_cutoff and -exp_cov parameters.

b. Velvet has many parameters that are explained in the Velvet Manual. A sample useful way to run Velvet is as following:

velveth working_directory 21 -shortPaired source-dir/Paired-RNA-seq_reads.fa

velvetg working_directory -read_trkg yes

2- Now run the SSP on the Velvet working_directory.

The following is the options used to control SSP:

Arguments

-dir_name

Sets the name of the directory in which SSP will write all of its output. It should be the same as the working_directory determined for Velvet in the previous step.

-read_length

The length of RNA-seq short reads.

-Kmer_length

Sets the length of kmer that is equal to the value set to –hash_length parameter in velveth.

-equation_threshold

Sets the length of interval needed to solve Interval Integer Linear Programming.

-clustering_threshold

Tell SSP the maximum distance between two contigs to cluster them together.

-LP_upper_bound

Tells SSP the maximum number of contigs in each locus to use linear programming for reconstructing isofoms in that locus.

-min_trans_length

Tells SSP the minimum length of transcripts to report as reconstructed ones.

An example command for running SSP :

SSP.sh – dir_name=working_directory – read_length =45 – Kmer_length =21 – equation_threshold =0.5 -clustering_threshold=1 -LP_upper_bound =20 -min_trans_lengt=100

Testing the installation

Download the Sample Data
In the SSP directory type:

./SSP.sh

The Program asks you to enter the required parameters to run the program, so enter the following values one by one

-dir_Name=path to Sample folder

-read_length=45

-Kmer_length=21

-equation_threshold=0.5

-clustering_threshold=1

-LP_upper_bound=17

-min_trans_length=100

You should see the following output:

ShortReadCount Started at Thu Sep 19 12:31:03 2013

ShortReadCount Finished at Thu Sep 19 12:31:03 2013

No_Short_In_Bin_Contig calc started at Thu Sep 19 12:31:03 2013

No_Short_In_Bin_Contig calc finished at Thu Sep 19 12:31:03 2013

Distance Calc Started at Thu Sep 19 12:31:03 2013

Distance Calc Finished at Thu Sep 19 12:31:03 2013

Reading Distance File ...

Clustering ...

Pruning Edges ... Done

Finding Connected Components ...

Component #1 Done

Component #2 Done

Component #3 Done

Component #4 Done

Saving Clusters ...

Locus 0: 8

Locus 1: 8

Locus 2: 1

Locus 3: 1

The Transfrags file in Sample folder contains the reconstructed transcripts by SSP.

SSP Development Group

Zhaleh Safikhani, University of Tehran
Mehdi Sadeghi, National Institute of Genetic Engineering and Biotechnology
Hamid Pezeshk, University of Tehran
Changiz Ezlahchi, Shahid Beheshti University

Contact Us

In the case that you have questions, suggestions or comments,

send email to safikhani@ut.ac.ir