Catalog

  • Introduction

  • Installation

  • History version

  • License

Introduction
SPHot is a sequence-based approach to predict hot spots in protein-RNA complexes using an ensemble classifier. For each residue, we use 10 features extracted from each 83 AAindex physicochemical descriptors along with relative accessible surface area(RASA) and BLOSUM62 matrix features to encode. Then, three algorithms, radial basis function (RBF)-based support vector machine (SVM), sigmoid-based SVM, and k-nearest neighbor algorithm (k-NN) are applied to construct the ensemble vote classifier.
NOTE: It should be noted that the hot spots and non-hot spots predicted by SPHot are not based on interaction interfaces, due to there is no structrtal information used in our model.
Any problems of installing or running this program please contact us.
Installation

REQUIREMENTS


-Linux(64bit)
SPHot 1.0 runs on Linux, we only tested it on Linux 64 bit.

-python 2.7
SPHot 1.0 depends on python(2.7), numpy(1.11.3), pandas(0.19.2) and scikit-learn(0.18.1). You can easily install the packages needed for SPHot by install Anaconda(Python 2.7 version).

Download the software(x86) from the Anaconda WWW site:

https://repo.continuum.io/archive/Anaconda2-4.3.1-Linux-x86_64.sh

Use the following steps to install: 
bash Anaconda2-5.1.0-Linux-x86_64.sh
When the installation is completed, you may need to add anaconda to your environment variable,
using the following command:
export PATH=/path_to_Anaconda/bin:$PATH
where Path_to_Anaconda is the absolute path of the Anaconda you installed.

-NetSurfP 1.0
Relative accessible surface area(RASA) are very useful for predicting hot spot, and SPHot also uses RASA as a feature. RASA can be calculated by NetSurfP-1.0, which predicts the surface accessibility and secondary structure of residues in amino acid sequences.

NetSurfP-1.0 is a property of the Center for Biological Sequence Analysis.
It may be downloaded only by special agreement.
For academic users there is a download site at:

http://www.cbs.dtu.dk/cgi-bin/nph-sw_request?netsurfp

NetSurfP-1.0 requires blastpgp-2.2.18 and NR70 database. The results of NetSurfP-1.0 may be slightly different if another version of BLASTPGP or NR BLAST database are used.

You can download blastpgp-2.2.18 and NR70 by the following WWW site:

ftp://ftp.ncbi.nlm.nih.gov/blast/executables/legacy.NOTSUPPORTED/2.2.18/blast-2.2.18-x64-linux.tar.gz
http://www.cbs.dtu.dk/services/NetSurfP-1.0/nr70_db.tar.gz

INSTALLATION

1. Untar the package: 
tar -xzvf SPHot-1.0-Linux.tar.gz
This will produce a directory 'SPHot-1.0'.

2.Edit the script 'run_predict.py' in the 'SPHot-1.0' directory.

Edit the following variable settings:
NetSurfP (line 11) : the full path to the 'netsurfp-1.0/netsurfp' script on your system;
MODEL (line 12) : the full path to the SPHot model file on your system, 
                  the default model is 'MODEL/model'
BLOSUM (line 13) : the full path to the BLOSUM62 file on your system, 
                   the default BLOSOM62 is 'data/blosum62'
AAindex1 (line 14) : the full path to the AAindex1(32 descriptors) file on your system, 
                     the default AAindex1 is 'data/AAindex1'
TMP (line 15) : the full path to the temporary directory of your choice 
                (must be user writable);
USAGE 

python run_predict.py [-h] [-i INFILE] [-s [SAVETMP]] [-t THRESHOLD] [-o OUT]
                      [-l [OUTFASTA]] [-n NETSURFP] [-m MODEL] [-B BLOSUM]
                      [-A AAINDEX1] [-p TEMPORARY]

Author: Leo   version: 1.0

optional arguments:
  -h, --help            show this help message and exit
  -i INFILE, --inFile INFILE
                        Input file name(must be fasta file)
  -s [SAVETMP], --savetmp [SAVETMP]
                        Keep all temporary directories[false]
  -t THRESHOLD, --threshold THRESHOLD
                        Predict a residue to be in a hot spot when 
                        the associated score is higher than threshold[0.578]
  -o OUT, --out OUT     Output file name
  -l [OUTFASTA], --outFasta [OUTFASTA]
                        Output in fasta format[false]
  -n NETSURFP, --NetSurfP NETSURFP
                        Path of NetSurfP1.0[netsurfp-1.0/netsurfp]
  -m MODEL, --model MODEL
                        Path of model file[MODEL/model]
  -B BLOSUM, --BLOSUM BLOSUM
                        Path of BLOSUM62 database[data/blosum62]
  -A AAINDEX1, --AAindex1 AAINDEX1
                        Path of AAindex1 database[data/AAindex1]
  -p TEMPORARY, --temporary TEMPORARY
                        Use this temporary directory[tmp]

History version