iCarPS

iCarPS provides a machine-learning method for carbonylation site prediction. It is implemented by WEKA library--random forest. At present, it provides user-friendly webserver to predict carbonylation sites freely, refering to http://lin-group.cn/server/iCarPS/webServer.html. Here we also provide our source codes for researchers who would like to run locally and conveniently to predict the query protein sequences in bulk.

Installation

Download iCarPS by

https://github.com/zhangdan0/iCarPS.git

Requirements

• Python>=3.0
• Java

Since the package is written in python 3.7, please use version 3.0 or above of python and the pip tool must be installed first. iCarPS uses the following dependencies: os, sys, re, numpy, pandas. You can install these packages first, by the following commands:

pip install os 
pip install sys
pip install re
pip install pandas 
pip install numpy

Besides, you must ensure that your local computer has a Java environment before running. You can open the cmd command window and input to check it:

java -version

Useage

• For users who want to perform carbonylation site prediction by our provided model :

  cd to the ./iCarPS_offline folder which contains iCarPS_offline.py and run as:

python iCarPS_offline.py [PredictType] [query sequences file]

1. the first parameter [PredictType] is used to assign a suitable type for prediction, including K, P, R, or T.
2. the second parameter [query sequences file] is the input sequence file, which must be fasta format. Please refer to those files in ./input/example folder.

Example:

python iCarPS_offline.py K ./input/test.txt

Finally, the result file is in output folder, named '*_finalresult.txt'