Project-Actaeon

Project Actaeon is a java wrapper for three weka models. The program can take membrane characteristics like the area per lipid and then predicts certain membrane components using machine learning algorithms.

Table of content

• Project description
• Installation
  • Prerequisites
  • Dependencies
• Arguments
  • Example command
• File specification
• Contact

Project description

This project is a java wrapper for three weka models. There are two RandomForest models and one LMT model. The two RandomForest model predict whether a sterol is present in the membrane and the type of tails in the lipids of the membrane. The LMT model predicts the sterol concentration in the membrane. You can find the process of building the models in the following git repo. The models use the following membrane characteristics: APL, thickness, bending rigidity, tilt angle, z-order and compressibility.

Installation

To install the program, simply pull this repo and use the following command to run the program. Of course, you will need to add your own arguments.

git pull https://github.com/Skippybal/Project-Actaeon.git

After pulling the git, you can now build the project using gradle and make a shadowJar. This jar will be created under build\libs\actaeon-1.0-SNAPSHOT-all.jar. To run the created jar use the following command (this works on Windows):

java -jar ./build/libs/actaeon-1.0-SNAPSHOT-all.jar [options]

Prerequisites

• Java version 16 SDK
• gradle

Dependencies

Group ID	Artifact ID	Version
commons-cli	commons-cli	1.4
nz.ac.waikato.cms.weka	weka-stable	3.8.0
commons-io	commons-io	2.6

Arguments

Parameter	Description	Type	Default
-f --infile	Input file, .arff or .csv	String
-o --outfile	Output file, .arff or .csv	String	data/output.csv
-p --print	Print the results to the command line		false
-a --apl	The apl of the membrane in nm^2	double
-t --thickness	The thickness of the membrane in nm	double
-b --bending	The apl of the membrane in kB T	double
-i --tilt	The tilt angle in degrees (°)	double
-z --zorder	The z-order of the membrane	double
-c --compress	The compressibility of the membrane in cN / m	double
-h --help	Print the help		false

IMPORTANT

• When an input file is specified, command line arguments like the apl and bending rigidity are ignored. This is to prevent overwriting issues in the output
• When no output file is specified, the output is written to data/output.csv
• When option -p is given, the results will still be written to the output file, but now also to the command line
• Options p and h both don't need a value, they are just given like -p and -h without a value

Example command

This is and example of a command that can be used to test the project. The command can be run after creating a shadowJar with gradle. This command will use the dat from the data/dataframe_all_sims3.csv file in the data folder from this repo. The results will be written to data/output.csv(default) and will also be written to the command line.

java -jar ./build/libs/actaeon-1.0-SNAPSHOT-all.jar -f data/dataFrame_all_sims3.csv -p

This is an example of how to use the program with input from the command line. Here the output is the same as the previous example.

java -jar ./build/libs/actaeon-1.0-SNAPSHOT-all.jar -a 0.736911 -t 3.628973 -b 10.0802 -i 14.7578 -z 0.175499 -c 24.18218 -p

File specifications

The input and output file both needs to be either .csv or .arff. The order of the columns in the input file is important.

In the case of .csv the input file must contain the following columns, in this specific order:

* APL
* thickness
* bending
* tilt
* zorder
* compress

In the case of .arff the input file must contain the following attributes, in this specific order:

* @attribute APL numeric
* @attribute thickness numeric
* @attribute bending numeric 
* @attribute tilt numeric 
* @attribute zorder numeric 
* @attribute compress numeric

IMPORTANT

• Rows with missing data are automatically removed from the input file.

Contact

• K. Notebomer
• [email protected]