Rescaling outputs from Cytosim

[ferrinm]

System and Software

• simulariumio Version: 1.12.0
• Python Version: 3.13.2
• Operating System: Ubuntu 23.04

Description

The scale_factor of MetaData within CytosimData does not behave as expected. When I set scale_factor=1.0 my data appears correctly oriented in the Simularium viewer, but setting any other value or leaving it to be automatically determined results in different objects being scaled differently from each other and possibly translated relatively to each other as well.

Expected Behavior

I expected setting a scale_factor > 1.0 to make all objects increase in scale by the same amount, so that everything visually looks the same in the viewer but I just don't have to zoom in as far to see it clearly.

Reproduction

Using the attached Cytosim outputs (point output_dir to the directory they are downloaded to in order to reproduce), run the Simularium conversion with scale_factor = 1.0 or 2.0. The first looks fine, the second has different objects scaled inconsistently.

singles.txt
couples.txt
solids.txt
fiber_points.txt

import os
import numpy as np
from simulariumio.cytosim import CytosimConverter, CytosimData, CytosimObjectInfo
from simulariumio import MetaData, DisplayData, DISPLAY_TYPE, ModelMetaData, InputFileData

output_dir = #PATH TO CYTOSIM OUTPUTS

for scale_factor in [1.0,2.0]:
    example_data = CytosimData(
        meta_data=MetaData(
            box_size=np.array([1.0,1.0,0.8]),
            scale_factor=scale_factor,
        ),
        object_info={
            "fibers" : CytosimObjectInfo(
                cytosim_file=InputFileData(
                    file_path=os.path.join(output_dir,"fiber_points.txt"),
                ),
                display_data={
                    0 : DisplayData(
                        name="actin",
                        radius=0.005,
                        display_type=DISPLAY_TYPE.FIBER
                    )
                },
                position_indices=[1, 2, 3]
            ),
            "solids" : CytosimObjectInfo(
                cytosim_file=InputFileData(
                    file_path=os.path.join(output_dir,"solids.txt"),
                ),
                display_data={
                    0 : DisplayData(
                        name="pit",
                        radius=0.04,
                        display_type=DISPLAY_TYPE.SPHERE
                    )
                },
                position_indices=[5, 6, 7]
            ),
            "singles" : CytosimObjectInfo(
                cytosim_file=InputFileData(
                    file_path=os.path.join(output_dir,"singles.txt"),
                ),
                display_data={
                    0 : DisplayData(
                        name="hip1r",
                        radius=0.005,
                        display_type=DISPLAY_TYPE.SPHERE
                    ),
                    1 : DisplayData(
                        name="linear actin nucleator",
                        radius=0.005,
                        display_type=DISPLAY_TYPE.SPHERE
                    ),
                    2 : DisplayData(
                        name="myosin",
                        radius=0.005,
                        display_type=DISPLAY_TYPE.SPHERE
                    )
                },
                position_indices=[3, 4, 5]
            ),
            "couples" : CytosimObjectInfo(
                cytosim_file=InputFileData(
                    file_path=os.path.join(output_dir,"couples.txt"),
                ),
                display_data={
                    0 : DisplayData(
                        name="arp2/3",
                        radius=0.005,
                        color="#bf95d4",
                        display_type=DISPLAY_TYPE.SPHERE
                    )
                },
                position_indices=[4, 5, 6]
            )
        },
    )

    CytosimConverter(example_data).save(os.path.join(output_dir,f'scale_factor={scale_factor}'))

[mattakamatsu]

(sorry I accidentally moved the issue and moved it back)

Once this bug is fixed, we'll use simularium in our social media post about Max's preprint 🤞🏼

[ascibisz]

Hello! The bug fix is now in, and it has been deployed to PyPi in the latest version of simulariumio, 1.12.1