compare_genes.md

Gene Content Dynamics

These scripts will allow you to compare the gene content of a species between all pairs of metagenomic samples.

Before running these scripts, you'll need to have run:
merge_midas.py genes read more.

Command usage:

compare_genes.py --indir <PATH> --out <PATH> [options]

  --indir PATH          Path to output from 'merge_midas.py genes' for one species
                        directory should be named according to a species_id and contains files 'genes_*.txt')
  --out PATH            Path to output file

Options:

  --max_genes INT       Maximum number of genes to use. Useful for quick tests (use all)
  --max_samples INT     Maximum number of samples to use. Useful for quick tests (use all)
  --distance {jaccard,euclidean,manhattan}
                        Metric to use for computing distances (jaccard)
  --dtype {presabs,copynum}
                        Data type to use for comparing genes (presabs)
  --cutoff FLOAT        Cutoff to use for determining presence absence (0.35)

Examples:

1. Run with defaults:
   compare_genes.py --indir /path/to/species --out distances.txt

2. Run a quick test:
   compare_genes.py --indir /path/to/species --out distances.txt --max_genes 1000 --max_samples 10

3. Use a different distance metric:
   compare_genes.py --indir /path/to/species --out distances.txt --distance manhattan

4. Use a lenient cutoff for determining gene presence absence:
   compare_genes.py --indir /path/to/species --out distances.txt --cutoff 0.10

5. Use a strict cutoff for determining gene presence absence:
   compare_genes.py --indir /path/to/species --out distances.txt --cutoff 0.75

Output format:

sample1: first sample identifier
sample2: second sample identifier
count1: number of present genes in sample1 count2: number of present genes in sample2
count_either: number of genes in sample1 or sample2
count_both: number of genes in sample1 and sample2
distance: dissimilarity between gene sets