Skip to content

Latest commit

 

History

History
55 lines (39 loc) · 2.65 KB

README.md

File metadata and controls

55 lines (39 loc) · 2.65 KB
Raw

Deviation_from_complete_graph

Author: Nolan K Newman [email protected] Date: 7/27/20 Created with Python v3.5.1

Purpose: Calculate how the observed network deviates from an ideal, complete graph that contains all expected correlations.

Description: This file takes as input a network file (format below) and will calculate all edge:node ratios (AKA "density"), PUC-compliant edges, and how the network deviates from a complete graph (i.e. all nodes connect to all other nodes) with the same number of nodes. As of right now, the structure of the network file it takes in is specific as it is the output of a previous file in a pipeline

Dependencies: none (all from base Python)

Example command: python dev_from_expected.py --input path/to/network/file --num_groups

Arguments: Required:

  • --input - network file of correlations between parameters, including directions of correlations, fold changes, and whether the edge is PUC-compliant
  • --num_groups - number of groups that correlations were originally calculated in. For example, if you calculated correlations in two different groups, WD and ND, then your argument would be "--num_groups 2"

Optional:

  • None at this time

Example input network header: pair,partner1,partner2,pval_E1,pval_E2,comb_pval,comb_rho,comb_FDR,partner1InFold,partner1_FoldChange,partner2InFold,partner2_FoldChange,corr_direction,partner1_FC_direction,partner2_FC_direction,IfFoldChangeDirectionMatch,PUC

  • pair: gene1 <==> gene2
  • partner1: gene1
  • partner2: gene2
  • pval_E1: correlation pvalue in experiment 1
  • pval_E2: correlation pvalue in Experiment 2
  • comb_pval: Fisher's combined pvalue across both experiments
  • comb_rho: combined rho coefficient across both experiments
  • comb_FDR: FDR calculated off the combined pvalue
  • partner1InFold: gene1
  • partner1_FoldChange: Fold change of gene1
  • partner2InFold: gene2
  • partner2_FoldChange: Fold change of gene2
  • corr_direction: correlation direction (either -1 or 1)
  • partner1_FC_direction: Fold change direction of gene1 (either -1 or 1)
  • partner2_FC_direction: Fold change direction of gene2 (either -1 or 1)
  • IfFoldChangeDirectionMatch: Are the previous 2 values identical (1 if yes, -1 if no)
  • PUC: Is the correlation PUC-compliant (are neg-neg or pos-pos correlations +ve and pos-neg correlations -ve?)

Output: Outputs a text file with each calculated parameter and its value. Values are rounded to 3 decimal places.

Notes: Will not work with a network that contains 0 negative edges, since you must divide the number of positive edges by the number of negative edges. Networks without negative edges are likely erroneous, anyways.