DIAMOnD

This is a guide on how to run the DIAMOnD algorithm, as described in Ghiassian et al. PLoS Comput Biol (2015).

Introduction

DIAMOnD is a network-based methodology to uncover the disease module associated with a particular phenotype. DIAMOnD uses as input the interactome and a set of proteins known to be associated with the phenotype of interest, which we call seeds. Then, it uses an iterative algorithm to calculate the significance of the interactions of the proteins in the neighborhood of the seeds (i.e. a p-value indicating if the number of interactions is higher than a random expectation). The proteins are ranked according to their respective p-values, and the protein with the highest rank (i.e. lowest p-value) is added to the set of seeds, increasing their number. These procedure is is repeated, increasing progressively the size of the disease module until a given number of steps is reached.

Python

• Version: [V1]
• Creator: Susan Dina Ghiassian, Joerg Menche & Albert-Laszlo Barabasi
• Last code update: 2015
• Last data update: ??
• Keywords: DIAMOnD; disease module identification; network medicine
• Rights Statement: ??

Instructions

1. Download the code, either by downloading the script code/DIAMOnD.py or by cloning the repository (git clone https://github.com/Barabasi-Lab/DIAMOnD.git).
2. Be sure to make the code executable by running the following command in the terminal: chmod +x DIAMOnD.py.
3. Run the code: ./DIAMOnD.py or python DIAMOnD.py in command line. Detailed instructions of how the code can be executed and an example can be found below.

Execution

DIAMOnD can be executed using the following commands:

./DIAMOnD.py <network_file> <seed_file> <n> <alpha>(optional) <outfile_name>(optional)

Or

python DIAMOnD.py <network_file> <seed_file> <n> <alpha>(optional) <outfile_name>(optional)

Where the parameters are:

• <network_file>: Path of the file of the network. The network must have the format of an edgelist: <node 1><separation><node 2>\n
• <seed_file>: Path of the file containing the seeds. Each seed must be separated by a new line: <seed 1>\n<seed 2>\n...<seed n>
• <n>: Desired number of steps. 200 is a reasonable starting point.
• <alpha> (optional): An integer representing the weight of the seeds. Default value is 1.
• <network_file> (optional): Path of the output file.

Requirements

• Python 2
• The following modules installed:
  • networkx
  • numpy
  • scipy.stats

Example

We can run the following example, using the input files in example/inputs

• seed_genes.txt: List of genes associated with a phenotype of interest.

2717
175
4669
2998
4125
...
51128
6609
2548
5476
6448

• PPI.txt: Example of protein-protein interaction network (note that identifiers in both input files should be consistent).

3920,5476
113457,4214
113457,7132
113457,1326
113457,1147
...
7316,6233
3423,3425
790,790
9513,9513
6233,6233

The following command runs the example:

./code/DIAMOnD.py example/inputs/PPI.txt example/inputs/seed_genes.txt 100 1 example/outputs/result.txt

It generates a list of the 100 non-seed nodes that have been ranked the highest by DIAMOnD in each iteration. The results are saved in the output file example/outputs/result.txt, which looks like this:

#rank	DIAMOnD_node
1	4758
2	3073
3	3074
4	3938
5	2799
...
96	5597
97	2316
98	4088
99	178
100	5601