Skip to content

Latest commit

 

History

History
226 lines (192 loc) · 7.54 KB

README.md

File metadata and controls

226 lines (192 loc) · 7.54 KB
Raw

Immpload - Immport upload preparation

immpload converts input data files into files formatted from Immport upload templates.

Prerequisites

  • Python3 with the pip package installer

Installation

  1. Install the immpload Python package and executable:

    pip install immpload

Usage

The simplest case copies input columns whose name matches the corresponding output Immport template column:

$ immpload subjectAnimals /path/to/input/subjects.txt

which will create the Immport upload file subjectAnimals.txt in the current directory.

To place the output in a different directory, use the -o or --outDir option:

$ immpload -o /path/to/output subjectAnimals /path/to/input/subjects.xslx

Note that the input can be either a .xslx Excel spreadsheet or a tab-delimited text file.

The command:

$ immpload --help

shows all immpload arguments and options.

Mapping Configuration

It is often useful to specify the conversion mapping in a YAML configuration file. For example, the following configuration:

columns:
    Subject ID: ID
    Arm Or Cohort ID: Cohort

converts the ID and Cohort input values to Subject ID and Arm Or Cohort ID output values, resp. The command is invoked with the -c or --config option, e.g:

$ immpload -o /path/to/output --config /path/to/conf/subjects.yaml \
           subjectAnimals /path/to/input/subjects.xslx

The configuration can include value mappings, e.g.:

values:
    Species: Mus musculus

sets the output Species to Mus musculus for all rows.

The configuration:

columns:
    Gender: Sex
values:
    Gender:
        n/a: Not Specified

transforms the input Sex value n/a to the output Gender value Not Specified. Other input values are copied without change.

immpload can flatten each input row into several output rows based on matching input column names against a pattern. For example, the configuration:

columns:
    Subject ID: ID
    Arm Or Cohort ID: Cohort
    Study Day: day
patterns:
    Result Value Reported: D(?P<day>\d+)$

converts an input row with columns D1, D2 and D3 into three output rows with column Study Day values 1, 2 and 3 and Result Value Reported values given by the D1, D2 and D3 input values, resp.

Immport upload data can be derived solely from fields embedded in column names. For example, the configuration:

columns:
    Analyte Reported: analyte
patterns:
    Analyte Reported: (?P<subject>.+)_(?P<day>.+)_(?P<analyte>.+)$

matches the input column names against the given pattern and writes one output row per matching column with the Analyte Reported column set to the embedded analyte match value. In this case, no other input rows are read besides the first header row of column names. Note that Analyte Reported is assigned the match value rather than the matching column value.

Defaults

immpload supplies certain required output columns with a reasonable default, as follows:

  • Animal Subjects (subjectAnimals.txt)

    • Age Unit - Days
    • Age Event - Age at infection

  • Experiment Samples (experimentSamples.*.txt)

    • Experiment ID - lower-case, underscored Experiment Name
    • Biosample ID - Expsample ID, if present, otherwise the lower-case, underscored Biosample Name, if present, otherwise derived from the Subject ID, Treatment ID and Experiment ID
    • Expsample ID - Biosample ID (defaulted, if necessary)

  • Treatments (treatments.txt)

    • Name - derived from the values and units
    • User Defined ID - lower-case, underscored Name
    • Use Treatment? - default is Yes

  • Assessments (assessments.txt)

    • Planned Visit ID - Study ID followed by d and the Study Day
    • Panel Name Reported - copied from the Assessment Type
    • Assessment Panel ID - derived from the Panel Name Reported
    • User Defined ID - derived from the Subject ID, Planned Visit ID and Component Name Reported

The default is set if and only if the mapped column value is missing.

Defaults are disabled with the --no-defaults option, e.g.:

$ immpload -o /path/to/output --config /path/to/conf/subjects.yaml \
           --no-defaults subjectAnimals /path/to/input/subjects.xslx

This is useful when submitting an update to an existing upload.

Validation

By default, imppload checks the output for required fields. If a required field is missing, then an error message is displayed and processing is halted.

Validation is disabled with the --no-validate option, e.g.:

$ immpload -o /path/to/output --config /path/to/conf/subjects.yaml \
           --no-validate subjectAnimals /path/to/input/subjects.xslx

As with no-defaults, no-validate is useful when submitting an update to an existing upload.

Callbacks

For advanced usage, the immpload Python module can be used directly in a Python script with a callback function, e.g.:

from immpload import munger

def add_results(in_row, in_col_ndx_map, out_col_ndx_map, out_row):
    """
    Modifies the output row after the configuration-based conversion.

    :param: in_row: the input data row
    :param: in_col_ndx_map: the input {column: index} dictionary
    :param: out_col_ndx_map: the output {column: index} dictionary
    :param: out_row :the output row
    :return: a list of rows derived from the given output row
    """
    ###
    ### Modify out_row or create new output rows here...
    ###
    # Return an array of rows.
    return [out_row]

# Convert the input file.
munger.munge('assessments', /path/to/input.xslx, callback=add_results)

The munger.munge method signature is as follows:

def munge(template, *in_files, config=None, out_dir=None,
          sheet=None, input_filter=None, callback=None, **kwargs):
    """
    Builds the Immport upload file for the given input file.
    The template is a supported Immport template name, e.g.
    `assessments`. The output is the Immport upload file,
    e.g. `assessments,txt`, placed in the output directory.

    The keyword arguments (_kwargs_) are static output
    _column_`=`_value_ definitions that are applied to every
    output row. The column name can be underscored, e.g.
    `Study_ID`.

    Output validation is disabled by default, but recommended
    for new uploads. Enable validation by setting the _validate_
    flag parameter to `True`.

    :param template: the required Immport template name
    :param in_files: the input file(s) to munge
    :param config: the configuration dictionary or file name
        of list of file names
    :param out_dir: the target location (default current directory)
    :param sheet: for an Excel workbook input file, the sheet to open
    :param input_filter: optional input row validator which has
        parameter in_row and returns whether the row is valid
    :param callback: optional callback with parameters
        in_row, in_col_ndx_map, out_col_ndx_map and out_row returning
        an array of rows to write to the output file
    :param defaults_opt: flag indicating whether to add defaults to the
        output (default `True`)
    :param validate_opt: flag indicating whether to validate the
        output for required fields (default `True`)
    :param append_opt: append rather than overwrite an existing output
        file (default False)
    :param kwargs: the optional static _column_`=`_value_ definitions
    :return: the output file name
    """