view track_objects.py @ 6:667a513e67e4 draft default tip

planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools commit 57a0433defa3cbc37ab34fbb0ebcfaeb680db8d5
author bgruening
date Sun, 05 Nov 2023 09:24:38 +0000
parents b178453ea8d1
children
line wrap: on
line source

#!/usr/bin/env python

import argparse
import json

from cp_common_functions import (get_json_value,
                                 get_pipeline_lines,
                                 get_total_number_of_modules,
                                 INDENTATION, update_module_count,
                                 write_pipeline)

MODULE_NAME = "TrackObjects"
OUTPUT_FILENAME = "output.cppipe"


def build_header(module_name, module_number):
    result = "|".join(
        [
            f"{module_name}:[module_num:{module_number}",
            "svn_version:\\'Unknown\\'",
            "variable_revision_number:7",
            "show_window:True",
            "notes:\\x5B\\'Track the embryos across images using the Overlap method\\x3A tracked objects are identified by the amount of frame-to-frame overlap. Save an image of embryos labeled with a unique number across time.\\'\\x5D",
            "batch_state:array(\\x5B\\x5D, dtype=uint8)",
            "enabled:True",
            "wants_pause:False]\n",
        ]
    )
    return result


def build_main_block(input_params):
    result = INDENTATION.join(
        [
            f"{INDENTATION}Choose a tracking method:{get_json_value(input_params,'con_tracking_method.tracking_method')}\n",
            f"Select the objects to track:{get_json_value(input_params,'object_to_track')}\n",
        ]
    )

    tracking_method = get_json_value(
        input_params, "con_tracking_method.tracking_method"
    )

    obj_measurement = "None"  # default value
    if tracking_method == "Measurements":
        measurement_category = get_json_value(
            input_params,
            "con_tracking_method.con_measurement_category.measurement_category",
        )
        measurement = get_json_value(
            input_params, "con_tracking_method.con_measurement_category.measurement"
        )

        if measurement_category == "Intensity" or measurement_category == "Location":
            img_measure = get_json_value(
                input_params, "con_tracking_method.con_measurement_category.img_measure"
            )
            obj_measurement = f"{measurement_category}_{measurement}_{img_measure}"
        else:
            obj_measurement = f"{measurement_category}_{measurement}"

    result += INDENTATION.join(
        [
            f"{INDENTATION}Select object measurement to use for tracking:{obj_measurement}\n"
        ]
    )

    if tracking_method == "LAP":  # no max distance required, set default for pipeline
        max_distance = 50
    else:
        max_distance = get_json_value(input_params, "con_tracking_method.max_distance")

    result += INDENTATION.join(
        [f"{INDENTATION}Maximum pixel distance to consider matches:{max_distance}\n"]
    )

    display_option = get_json_value(input_params, "con_tracking_method.display_option")

    output_img_name = "TrackedCells"  # default value, required by cppipe regardless of its presence in UI
    save = get_json_value(
        input_params, "con_tracking_method.con_save_coded_img.save_coded_img"
    )
    if save == "Yes":
        output_img_name = get_json_value(
            input_params, "con_tracking_method.con_save_coded_img.name_output_img"
        )

    result += INDENTATION.join(
        [
            f"{INDENTATION}Select display option:{display_option}\n",
            f"Save color-coded image?:{save}\n",
            f"Name the output image:{output_img_name}\n",
        ]
    )

    # LAP method default values
    movement_model = "Both"
    no_std = 3.0
    radius_limit_max = 10.0
    radius_limit_min = 2.0
    radius = "2.0,10.0"
    run_second = "Yes"
    gap_closing = 40
    split_alt = 40
    merge_alt = 40
    max_gap_displacement = 5
    max_split = 50
    max_merge = 50
    max_temporal = 5
    max_mitosis_dist = 40
    mitosis_alt = 80

    # LAP method
    if tracking_method == "LAP":
        movement_model = get_json_value(
            input_params, "con_tracking_method.movement_method"
        )
        no_std = get_json_value(input_params, "con_tracking_method.no_std_radius")
        radius_limit_max = get_json_value(
            input_params, "con_tracking_method.max_radius"
        )
        radius_limit_min = get_json_value(
            input_params, "con_tracking_method.min_radius"
        )
        radius = f"{radius_limit_min},{radius_limit_max}"

        run_second = get_json_value(
            input_params, "con_tracking_method.con_second_lap.second_lap"
        )
        if run_second == "Yes":
            gap_closing = get_json_value(
                input_params, "con_tracking_method.con_second_lap.gap_closing"
            )
            split_alt = get_json_value(
                input_params, "con_tracking_method.con_second_lap.split_alt"
            )
            merge_alt = get_json_value(
                input_params, "con_tracking_method.con_second_lap.merge_alt"
            )
            max_gap_displacement = get_json_value(
                input_params, "con_tracking_method.con_second_lap.max_gap_displacement"
            )
            max_split = get_json_value(
                input_params, "con_tracking_method.con_second_lap.max_split"
            )
            max_merge = get_json_value(
                input_params, "con_tracking_method.con_second_lap.max_merge"
            )
            max_temporal = get_json_value(
                input_params, "con_tracking_method.con_second_lap.max_temporal"
            )
            max_mitosis_dist = get_json_value(
                input_params, "con_tracking_method.con_second_lap.max_mitosis_distance"
            )
            mitosis_alt = get_json_value(
                input_params, "con_tracking_method.con_second_lap.mitosis_alt"
            )

    result += INDENTATION.join(
        [
            f"{INDENTATION}Select the movement model:{movement_model}\n",
            f"Number of standard deviations for search radius:{no_std}\n",
            f"Search radius limit, in pixel units (Min,Max):{radius}\n",
            f"Run the second phase of the LAP algorithm?:{run_second}\n",
            f"Gap closing cost:{gap_closing}\n",
            f"Split alternative cost:{split_alt}\n",
            f"Merge alternative cost:{merge_alt}\n",
            f"Maximum gap displacement, in pixel units:{max_gap_displacement}\n",
            f"Maximum split score:{max_split}\n",
            f"Maximum merge score:{max_merge}\n",
            f"Maximum temporal gap, in frames:{max_temporal}\n",
        ]
    )

    # common section
    filter_by_lifetime = get_json_value(
        input_params, "con_tracking_method.con_filter_by_lifetime.filter_by_lifetime"
    )
    use_min = "Yes"  # default
    min_life = 1  # default
    use_max = "No"  # default
    max_life = 100  # default

    if filter_by_lifetime == "Yes":
        use_min = get_json_value(
            input_params,
            "con_tracking_method.con_filter_by_lifetime.con_use_min.use_min",
        )
        if use_min == "Yes":
            min_life = get_json_value(
                input_params,
                "con_tracking_method.con_filter_by_lifetime.con_use_min.min_lifetime",
            )

        use_max = get_json_value(
            input_params,
            "con_tracking_method.con_filter_by_lifetime.con_use_max.use_max",
        )
        if use_max == "Yes":
            max_life = get_json_value(
                input_params,
                "con_tracking_method.con_filter_by_lifetime.con_use_max.max_lifetime",
            )

    result += INDENTATION.join(
        [
            f"{INDENTATION}Filter objects by lifetime?:{filter_by_lifetime}\n",
            f"Filter using a minimum lifetime?:{use_min}\n",
            f"Minimum lifetime:{min_life}\n",
            f"Filter using a maximum lifetime?:{use_max}\n",
            f"Maximum lifetime:{max_life}\n",
        ]
    )

    # print 2 leftover from LAP
    result += INDENTATION.join(
        [
            f"{INDENTATION}Mitosis alternative cost:{mitosis_alt}\n",
            f"Maximum mitosis distance, in pixel units:{max_mitosis_dist}\n",
        ]
    )

    # Follow Neighbors
    # defaults
    avg_cell_diameter = 35.0
    use_adv = "No"
    cost_of_cell = 15.0
    weight_of_area_diff = 25.0

    if tracking_method == "Follow Neighbors":
        avg_cell_diameter = get_json_value(
            input_params, "con_tracking_method.avg_diameter"
        )
        use_adv = get_json_value(
            input_params, "con_tracking_method.con_adv_parameter.adv_parameter"
        )
        if use_adv == "Yes":
            cost_of_cell = get_json_value(
                input_params, "con_tracking_method.con_adv_parameter.cost"
            )
            weight_of_area_diff = get_json_value(
                input_params, "con_tracking_method.con_adv_parameter.weight"
            )

    result += INDENTATION.join(
        [
            f"{INDENTATION}Average cell diameter in pixels:{avg_cell_diameter}\n",
            f"Use advanced configuration parameters:{use_adv}\n",
            f"Cost of cell to empty matching:{cost_of_cell}\n",
            f"Weight of area difference in function matching cost:{weight_of_area_diff}\n",
        ]
    )
    result = result.rstrip("\n")
    return result


if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("-p", "--pipeline", help="CellProfiler pipeline")
    parser.add_argument("-i", "--inputs", help="JSON inputs from Galaxy")
    args = parser.parse_args()

    pipeline_lines = get_pipeline_lines(args.pipeline)
    inputs_galaxy = json.load(open(args.inputs, "r"))

    current_module_num = get_total_number_of_modules(pipeline_lines)
    current_module_num += 1
    pipeline_lines = update_module_count(pipeline_lines, current_module_num)

    header_block = build_header(MODULE_NAME, current_module_num)
    main_block = build_main_block(inputs_galaxy)

    module_pipeline = f"\n{header_block}{main_block}\n"
    pipeline_lines.append(module_pipeline)

    write_pipeline(OUTPUT_FILENAME, pipeline_lines)