/DESMOND

Primary LanguageJupyter Notebook

DESMOND

DESMOND is a method for identification of Differentially ExpreSsed gene MOdules iN Diseases.

DESMOND accepts gene expression matrix and gene interaction network and identifies connected sets of genes up- or down-regulated in subsets of samples.

alt text

Input

  • matrix of normalized gene expressions; first row and column contain gene and sample names respectively
  • network of gene interactions; list of edges in a format gene1\tgene2 or network in NDEx format https://home.ndexbio.org/about-ndex/

Usage example

python DESMOND.py --exprs $exprs --network $network  --basename $proj_name --out_dir $outdir \
 --alpha $a --p_val $p_val -q $q  --direction UP --verbose  > $outdir/$proj_name.UP.LOG 2> $outdir/$proj_name.UP.ERR;
python DESMOND.py --exprs $exprs --network $network  --basename $proj_name --out_dir $outdir \
 --alpha $a --p_val $p_val -q $q  --direction DOWN --verbose > $outdir/$proj_name.DOWN.LOG 2> $outdir/$proj_name.DOWN.ERR;

# calculate empirical p-values and merge up- and dow-regulated biclusters if necessary

python post-processing.py --up $outdir/$proj_name.'alpha='$a',beta_K='$b',direction=UP,p_val='$p_val',q='$q.biclusters.tsv \
--down $outdir/$proj_name.'alpha='$a',beta_K='$b',direction=DOWN,p_val='$p_val',q='$q.biclusters.tsv \
--exprs $exprs --network $network -s $outdir/$proj_name',q='$q'.SNR_threshold.txt' \
--out $outdir/$proj_name.'alpha='$a',beta_K='$b',p_val='$p_val',q='$q.biclusters.permutations.tsv  --verbose  > $outdir/$proj_name.permutations.LOG 2> $outdir/$proj_name.permutations.ERR;

Output

  • *.biclusters.tsv - list of identified biclusters.
  • *.network.txt - temporary network file, contains the network with samples assigned on edges. This file is used for restarts with the same network and parameters 'direction', 'p_val', 'min_SNR' and 'min_n_samples'.

Cite

code used in the paper: DESMOND_py2.zip

Zolotareva et al. Identification Of Differentially Expressed Gene Modules In Heterogeneous Diseases. (2020) Bioinformatics.

@article{Zolotareva2020,
  doi = {10.1093/bioinformatics/btaa1038},
  url = {https://doi.org/10.1093/bioinformatics/btaa1038},
  year = {2020},
  month = dec,
  publisher = {Oxford University Press ({OUP})},
  author = {Olga Zolotareva and Sahand Khakabimamaghani and Olga I Isaeva and Zoe Chervontseva and Alexey Savchik and Martin Ester},
  editor = {Janet Kelso},
  title = {Identification Of Differentially Expressed Gene Modules In Heterogeneous Diseases},
  journal = {Bioinformatics}
}