Package ubic.gemma.core.loader.expression.simple

Interface ExperimentalDesignImporter

All Known Implementing Classes:

ExperimentalDesignImporterImpl

public interface ExperimentalDesignImporter

Parse a description of ExperimentalFactors from a file, and associate it with a given ExpressionExperiment. The format is specified by ExperimentalDesignWriter

Example of format, where 'Category' is an MGED term and 'Type' is either "Categorical' or 'Continuous' with no extra white space around the '='s. The ID column MUST match the names on the BioAssays the design will be attached to. Main section is tab-delimited. Column headings in the main table must match the identifiers given in the header.

    #$ Age : Category=Age Type=Continuous
    #$ Profile : Category=DiseaseState Type=Categorical
    #$ PMI (h) : Category=EnvironmentalHistory Type=Continuous
    #$ Lifetime Alcohol : Category=EnvironmentalHistory Type=Categorical
    #ID  Age     Profile     PMI (h)     Lifetime Alcohol
    f-aa     50  Bipolar     48  Moderate present
    f-ab     50  Bipolar     60  Heavy in present
    f-ac     55  Schizophrenia   26  Little or none
    f-ad     35  Bipolar     28  Unknown
    f-af     60  Bipolar     70  Little or none
 

Note: Files downloaded from Gemma may have an "ExternalID" column after the first column. This is allowed in the import format.

Author:

Paul

See Also:

ExperimentalDesignWriter

Method Summary

Modifier and Type	Method	Description
void	importDesign(ExpressionExperiment experiment, InputStream is)	This is the main builder director method of the application: It processes the input file containing information about the experimental design for a given expression experiment.

Method Detail

importDesign

@Secured({"GROUP_USER","ACL_SECURABLE_EDIT"})
void importDesign(ExpressionExperiment experiment,
                  InputStream is)
           throws IOException

This is the main builder director method of the application: It processes the input file containing information about the experimental design for a given expression experiment. There are 3 main steps in the workflow:

Step1 - validate the the 3 components of the file which should contain: The first component is the experimental factor lines marked with a $# there are as many lines as experimental factors - (experimentalFactorLines). Then one line containing header information indicating the order of the experimental factors in the file - sampleHeaderLine Finally factor values, the first column of which is the sample or biomaterial ids and thereafter factor values.

Step 2 the file components are mapped to objects to populate the experimental design, before addition of objects to the composite existing values are checked for. The expression experiment composite: ExpressionExperiments have an experimental design which have experimental factors. Experimental factors have factor values. BioMaterials have factor values. Bioassays have biomaterials, bioassays are in an expression experiment which completes the circle.

Step 3 on successful validation and object creation the experimental design is persisted following a strict order, expression factors first then biomaterial details.

Parameters:

experiment - the one to add the experimental design

is - File to process

Throws:

IOException - when IO problems occur.

See Also:

#importDesign(ubic.gemma.model.expression.experiment .ExpressionExperiment, java.io.InputStream, boolean)