Skip to content

Step 1 (S1): Generate semantic segmentations

Running S1

1. Denoise tomograms

First, denoise the input tomograms using TomoEED by running the following command:

<tomoeed_directory>/bin/tomoeed full_path_to_input_tomogram.mrc denoised_tomograms/output_tomogram.mrc
This will make a denoised version of the input tomogram (input_tomogram.mrc) and place it at denoised_tomograms/output_tomogram.mrc. If running from a different directory than input_tomogram.mrc, you will need to specify the full path to the input tomogram.

Important

Make sure to use the path to denoised tomogram (denoised_tomograms/output_tomogram.mrc) in the s1_param_file_path for tomogram in the inputs section.

2. Generate semantic segmentations

Run the following command to generate the semantic segmentations.

python src/picket/s1.py <s1_param_file_path>
Next step is to choose one or more optimal segmentations for each input tomogram and obtaining the corresponding particle cluster IDs. Follow the instructions at visualizing segmentations and obtaining particle cluster ID for more details on this.



Back to Home
Go to outputs
Go to usage instructions