Skip to content

thomasgaudelet/hypergraphlets

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Hypergraphlets counter

Implementation of hypergraphlet counter used in

Gaudelet, T., Malod-Dognin, N. and Pržulj, N., 2018. Higher-order molecular organization as a source of biological function. Bioinformatics, 34(17), pp.i944-i953.

It is now updated to output both the hypergraphlets and the simplets counts. Simplets were introduced in

Malod-Dognin, N. and Pržulj, N., 2019. Functional geometry of protein interactomes. Bioinformatics, 35(19), pp.3727-3734.

Both hypergraphlets and simplets give features to characterize the wiring patterns around nodes in hypergraphs (or simplicial complexes). The simplets are a restriction of hypergraphlets and give a smaller signature. We refer the reader to both papers above for more information.

The implementation requires BOOST library (https://www.boost.org/) and a compiler supporting C++11.

Before compiling add the absolute path to the BOOST lib/ folder on your system to the Makefile Compile by opening a terminal, navigating to the hypergraphlets folder, and entering the command make.

To run the counter simply enter the command

./run_hypercounter  -g  "path_to_file/hyperedge.list"  -o  "path_to_output/name_of_output"  -t  "Number of threads, default 1"  -b  "Number of groups of genes to launch across the threads, default 50"

The hyperedge list should follow the format: - tab separated - each row starts with the hyperedge index (between 0 and m-1, where m is the total number of hyperedges) and contain the list of vertices it contains - vertices should be indexed from 0 to n-1, n being the total number of vertices

The counter outputs two .svml files containing the hypergraphlets and simplets counts with the following format: - each row correspond to the count for a vertex - the first number (and last) correspond to the vertex index - the count for each orbit is of the shape orbit:count

The code uses OpenMP for parallelisation, the number of threads to use can be set using the flag -t.

The vertex sets is partitioned in a number of blocks (set by the flag -b) each block being processed by a single thread. The partition is randomised to avoid having consecutive vertex in the same block (this is to avoid having the densely connected vertices in the same block)

About

Hypergraphlet counter

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 100.0%