Differential modification analysis

Identify reference sites with a significant difference in modification stoichiometry between conditions.

Running diff test

Use SWARM_diff.py to run GLM test and identify reference coordinates where stoichiometry depends on the condition.

python3 SWARM_diff.py -i $INPUT -o $OUT -n 24

Required:
  -i, --data_file            Tab-separated file with columns: M2_file_path RepName Condition
  -o, --output_file          Output file path

  Optional:
  -p, --M2_threshold         Only test sites where 1+ sample has site probability over p [0]
  -s, --Stoich_threshold     Only test sites where 1+ sample has stoichiometry over s [0]
  -delta, --Delta_threshold  BH adjust for sites over minimum delta stoichiometry [0]
  -n, --ncpus                Number of CPU threads [1]
  -h, --help                 Show this help message and exit

Input format

Create an input file in tsv format, make sure to include header with 3 columns as seen below.

M2_file_path    RepName Condition
/path/WT_rep1.m2.pred.tsv   1   WT
/path/WT_rep2.m2.pred.tsv   2   WT
/path/KO_rep1.m2.pred.tsv   1   KO
/path/KO_rep2.m2.pred.tsv   2   KO

1. M2_file_path: path to the site-level prediction file (tsv)
2. RepName: name of the rep, each sample of the same condition requires unique name
3. Condition: reps will be grouped by condition, label the same condition consistently

Output format

Produces a tsv file with p-values and average and per-rep information for sites tested with enough coverage in at least one sample.

Visualization