Catfish implements an efficient heuristic to decompose a given flow into a set of minimum number of paths.
Please refer to our paper at here. The tools, scripts, and datasets used in this manuscript can be found at catfishtest.

Latest release of Catfish is v0.2.1, including both source code, and linux binary and macOS binary.

Download Catfish from source code. To install it, you need to first download Boost if it has not been installed in your system, and then compile the source code of Catfish.

If Boost has not been downloaded/installed in your system, download Boost (license) from (http://www.boost.org). Uncompress it somewhere (compiling and installing are not necessary).

Use the following to compile and install Catfish:

./configure --prefix=/path/to/your/installation/directory --with-boost=/path/to/your/boost/directory
make
make install

The --prefix argument for configure specifies the directory where you would put the binary of Catfish. It is optional and the default is /usr/local/bin for most linux distributions. If Boost has been installed in your system, the --with-boost argument for configure can also be omitted.

The usage of catfish is:

catfish -i <input.sgr> -o <output-file> [-a core|full|greedy] [-h] [-v]

-i parameter specifies the input file, which specifies multiple directed acyclic graph. Each graph starts with a header line marked with #. The next line gives n, indicating the number of vertices in this graph. The vertices are named from 0 to (n - 1), where vertex 0 has to be the source vertex and vertex (n - 1) has to be the sink vertex. Each of the following line specifies an edge, which consists of three integers: the in-vertex, out-vertex and the weight of this edge. There is one such input example files at examples.

-o parameter specifies the output file, which will show the predicted paths and their associated abundances.

-a parameter specifies the algorithm. There are three options: {full, core, greedy}. With option of core, the program will only run the core algorithm to partly decompose the given splice graph, which will predict fewer paths but with higher accuracy. With option of full, the program will completely decompose the given splice graph, using greedy algorithm following the core algorithm. With option of greedy, the program will only use greedy algorithm to fully decompose the given splice graph. This parameter is optional, and its default value is full.

-h parameter prints the usage of Catfish and exits.

-v parameter prints the version of Catfish and exits.