– Runs genome evaluations on PICRUSt.


Using files created by it runs each test dataset through the ASR ( and the genome prediction (

Usage: [options]

Input Arguments:



-i, --input_dir
Directory containing one or more test datasets
-t, --ref_tree
Reference tree that was used with make_test_datasets


-o, --output_dir
The output directory [default: <input_dir>]
-j, --parallel_method
Method for parallelization. Valid choices are: sge, torque, multithreaded [default: multithreaded]
-m, --prediction_method
Method for trait prediction. See for full documentation. Valid choices are: asr_and_weighting, nearest_neighbor, random_neighbor [default: asr_and_weighting]
If set, calculate confidence intervals with ace_ml or ace_reml, and use confidence intervals in trait prediction
If set, calculate accuracy using the NSTI (nearest sequenced taxon index) during trait prediction
-a, --asr_method
Method for ancestral_state_reconstruction. See for full documentation. Valid choices are: ace_ml, ace_reml, ace_pic, wagner [default: wagner]
-w, --weighting_method
Method for weighting during trait prediction. See for full documentation. Valid choices are: linear, exponential, equal [default: exponential]
-n, --num_jobs
Number of jobs to be submitted (if –parallel). [default: 100]
Location to store intermediate files [default: <output_dir>]
Run all jobs even if output files exist [default: False]
Check if pre-existing output files have null files. If so remove them and re-run. [default: False]


Predictions from for each test dataset.

Minimum Requirments:

Provide a directory that contains one or more datasets created by and the original reference tree used -i test_datasets_dir -t reference_tree_fp

Specify output file: -i test_datasets_dir -t reference_tree_fp -o output_dir

Force the launching of jobs that alredy seem done by overwriting existing output files: --force -i test_datasets_dir -t reference_tree_fp -o output_dir