This C++ code was programed to transfer a Species-Centric Network (SCNs, or species-species co-occurrence networks) to its corresponding Correlation-Centric Network (CCN,or edge network) in microbial community. In SCN, a node represents a species and an edge represents the correlation between the linked species. each node of a CCN represents a species-species co-occurrence, and each edge of CCN represents a species shared between two co-occurrences.
The mathematical derivation process of SCN are referred by the paper by T.S.Evans and R.Lambiotte, Line Graphs, Link Partitions and Overlapping Communities, Phys.Rev.E 80 (2009) 016105 arXiv:0903.2181.
These instructions will get you a copy of the project up and running on your local machine for transferring SCNs to corresponding CCNs in microbial community.
This software could be installed in both Linux system and Windows system. No dependent software needs to be installed
Basic Compiling You can try the following:
g++ -o correlationcenticnetwork TseGraph.cpp main.cpp
The TseGraph.o and main.o files can be deleted afterwards.
The g++.exe depends on your compiler, here the standard gnu c++ compiler is used. Below for the intel compiler on the altix icpc is needed. This produces an executable correlationcenticnetwork (on Windows its correlationcenticnetwork.exe). This is run by typing ./correlationcenticnetwork on Unix (simply correlationcenticnetwork on Windows). with any options you require. If you want to get more complicated, try optimizing options:
g++ -c -O2 -o TseGraph.o TseGraph.cpp g++ -c -O2 -o main.o main.cpp g++ -o correlationcenticnetwork TseGraph.o main.o
makefile
If you have make then the following makefile might work.
The makefile is stored as makefileTSE but needs to be renamed to be simply makefile.
Then just move to the directory with the main.cpp, TseGrapgcpp.cpp, TseGrapgcpp.h and makefile then type make.
For information on make try mrbook.org/tutorials/make/.
CC=g++
CFLAGS=-c -Wall
LDFLAGS=-O2
all: correlationcenticnetwork
correlationcenticnetwork: main.o TseGraph.o
$(CC) $(LDFLAGS) main.o TseGraph.o -o correlationcenticnetwork
main.o: main.cpp
$(CC) $(CFLAGS) main.cpp
factorial.o: factorial.cpp
$(CC) $(CFLAGS) TseGraph.cpp
clean:
rm -rf *o correlationcenticnetwork
To make a CCN type
correlationcenticnetwork -i input_file -o output_file [options]
where the options are: -t n : create CCN of type number n. Default is 2. -w : read the SCN as a weighted one, otherwise it is unweighted. -d : read the SCN as a directed one, otherwise it is undirected. -h : show this usage message. The CCN types are: 0: Unweighted and no self loops 1: Unweighted with self-loops 2: Weighted and no self loops 3: Weighted with self-loops
For example try:
correlationcenticnetwork -i ./example/14_dry.txt -o ./14dry_linegraph -t 3
The output file and the speices name of members of SCN is also provided in example file.
The input file is a two-column table like:
0 1
3 4
0 15
... ...
Where the first column is the source vertex, the second is the target vertex and the last is the edge weight.
Each number represents one species, and the correlation between species and numbers are stored in file 14dry_name.txt
For the output file
The input file is a three-column table like:
0 1 0.140351
0 2 0
0 3 0.157895
0 4 0.0619195
... ...
Where the first column is the source vertex, the second is the target vertex and the last is the edge weight.
The third line is the predicted correlation coefficient
Note that the vertices in the line graph are again numbered from 0 in the order of the edges in the input file.
This project is licensed under the GNU General Public License v3.0