MSIGNET Files

MSIGNET
===========
MSIGNET integrates disease-specific gene expression data and human protein-protein interactions in a Bayesian network, and identifies interactions of genes significantly expressed under the disease condition.

Reference:
Xi Chen, Jianhua Xuan. "MSIGNET: a Bayesian approach for disease-associated gene network identification", 2020.




Installation 
=============
MSIGNET was implemented in MATLAB and the script has been tested on both Windows and Linux systems. It can be ran using MATLAB version 2016a and later. 




Input
============= 
Gene_exp: Gene expression data (log2 based) for N genes and M samples (N x M matrix, M = M_d+M_c)

Gene_entrez_id: Gene Entrez ID for N genes (N x 1 vector)

early_idex (disease_idex): Indexes for disease samples or early recurrent samples specific cancer recurrence analysis (M_d x 1 vector)

late_idex (disease_idex): Indexes for control samples or late recurrent samples specific cancer recurrence analysis (M_c x 1 vector)

PPI_network: Protein-protein interactions for genes of interest (N_p x N_p sparse binary matrix, N_p < N)  

PPI_gene_id: Gene Entrez ID for N_p network genes (N_p x 1 vector)  

PPI_edge_zscore: Gene co-expression z-score for Pearson correlation coefficient for network connected genes (N_p x N_p sparse double matrix)  

PPI_gene_zscore: Gene differential expression z-score (based on t-statistics p-value) (N_p x 1 vector) 

Max_loop_Num: The number of sampling rounds for MCMC process (constant integer)

save_iterative_Num: Saving sampling results after a number of sampling rounds (constant integer)




Output
=============
PPI_gene_sampling_freqency: Sampling frequency for each of N_p genes (N_p x 1 vector)
 
PPI_edge_sampling_freqency: Sampling frequency for PPI connected gene pairs (N_p x N_p sparse matrix)



Input data for simulation, Parkinson's disease and ovarian cancer as discussed in the MSIGNET paper was provided together with the MSIGNET scripts.



Any feedback, comments and questions are always welcome. Please address them to Xi Chen (xichen86@vt.edu).