A datapackage representation of cell migration-derived tracking files.

This Python project aims to create a simple Python package to produce data packages of cell migration tracking files. The final goal is to have a uniform, standardized way to represent these data, as in Frictionless Data and Data Packages .

Steps to follow to use the package:

• step 1 - Install the package (note it's Python 3 only at the moment):

python setup.py install

• step 2 - create a biotracks.ini configuration file and place it in the same directory as your tracking file. The file must be structured as follows:

[TOP_LEVEL_INFO]
author = the author of the dp
title = a title describing the dp
name = a name for the dp
author_institute = the insitute of the author
author_email = a valid email address

[TRACKING_DATA]
x_coord_cmso = the column name pointing to the x coordinate
y_coord_cmso = the column name pointing to the y coordinate
z_coord_cmso = the column name pointing to the z coordinate
frame_cmso = the column name pointing to the frame information
object_id_cmso = the object identifier
link_id_cmso = the link identifier

• step 3 - move to the scripts directory and run:

python create_dpkg.py your_tracking_file

this will create a dp directory containing:

• a csv file for the objects (i.e. cells)
• a csv file for the links (i.e. a linear collection of objects)
• and a dp.json file containing the json schemas of the two csv files.

This last file will look something like this:

{
    "resources": [{
        "name": "objects_table",
        "schema": {
            "primaryKey": "SPOT_ID",
            "fields": [{
                "name": "SPOT_ID",
                "title": "",
                "description": "",
                "constraints": {
                    "unique": true
                },
                "type": "integer",
                "format": "default"
            }, {
                "type": "integer",
                "name": "FRAME",
                "title": "",
                "format": "default",
                "description": ""
            }, {
                "type": "number",
                "name": "POSITION_X",
                "title": "",
                "format": "default",
                "description": ""
            }, {
                "type": "number",
                "name": "POSITION_Y",
                "title": "",
                "format": "default",
                "description": ""
            }]
        },
        "path": "objects.csv"
    }, {
        "name": "links_table",
        "schema": {
            "foreignKeys": [{
                "fields": "SPOT_ID",
                "reference": {
                    "resource": "objects_table",
                    "fields": "SPOT_ID",
                    "datapackage": ""
                }
            }],
            "fields": [{
                "type": "integer",
                "name": "LINK_ID",
                "title": "",
                "format": "default",
                "description": ""
            }, {
                "type": "integer",
                "name": "SPOT_ID",
                "title": "",
                "format": "default",
                "description": ""
            }]
        },
        "path": "links.csv"
    }],
    "name": "CMSO_tracks",
    "title": "A CMSO data package representation of cell tracking data",
    "author_email": "paola.masuzzo@email.com",
    "author_institute": "VIB",
    "author": "paola masuzzo"
}

Then, the datapackage is pushed to a pandas dataframe. At the moment, this dataframe is used to create simple visualizations of links and turning angles.