Minimal Information About Microarray Experiments (MIAME): Definitions and relationships incorporating the Microarray Markup Language (MAML).

MIAME version March 21, 2001

MAML version January 19, 2001

 

 

Description

MAML Data format

When applicable

Dependencies

Allowed values

Experimental Design Concepts

Overview of the experimental study.

 

 

 

 

Author, laboratory, and links

Submitter and laborarory name of the MIAME. Also any associated URLs

contact

Always

Referenced by creation_info

Contact element attribute list

Contact information

Name, address, phone, FAX, email of person to contact. May be different from submitter

contact

Always

Referenced by experiment_set?

Contact element attribute list

Experiment type

The general kind of problem that the experiments are designed to  address.

Experiment_set attributes: type and other_type.

Handle more than one?

Always

Should be consistent with experimental factors.

experiment_set: type =

time_course, dose_response, single_hybridisation, disease_vs_normal, tissue_comparison, other (treated_vs_untreated, shock, gene_knock-out, gene_knock-in)

Experimental factors

Factor(s) that is varied between experiments. May be same as experiment type but need not be.

Experimental_factors

Always

Referenced by experiment_design which in turn is referenced by experiment_set

One or more text terms.

time, dose, genetic variation, response to treatment or compound, severity, location

MGED controlled vocab?

Platforms used

The type(s) of array(s) used.

possibly contained in array_definition: description

Always

Should be consistent with Array Design concepts.

In-house, commercial;

filter array (macroarray, membranes),

lithographic oligo microarray (Affymetrix),

cDNA microarray (attached to glass slides; 2-color fluorescence),

oligo array (attached to slide after synthesis; 2-color fluorescence)

MGED controlled vocab?

 Single or multiple hybridisations

Establish whether relationships between hybridisations are needed. Note that one 2-color array experiment is a single hybridisation.

Inferred from hybridisation_list

Always

Should be consistent with experiment type. I.e., not single if type = time_course or dose_response

Single, multiple

Ordered/ unordered (what if both? Describe separately?)

The order of the data from the multiple hybridisations is important.

Inferred from experiment_set: type

possibly contained in biology_description

Multiple hybridisations

Must be multiple hybridisation

Ordered, unordered

Serial

The hybridisations were done using sources sampled at successive regular intervals.

possibly contained in biology_description

Multiple hybridisations

Ordered

Yes, no

Serial type

The meaning of the serial nature of the hybridisations.

Experiment_set attributes: type and other_type.

Serial = yes

Should be consistent with experiment type.

Time_course, dose_response

Grouping

The hybridisations can be divided up into meaningful groups

None

Multiple hybridisations

Unordered (apply to ordered data?)

Yes, no

Grouping type

The meaning of the groups.

Experiment_set attributes: type and other_type.

Grouping = yes

Should be consistent with experiment type.

treatment, disease, knock-out, knock-in, shock, multiple tissue comparison, other

Sample and arrays list

The relationships of each sample and array to each other.

Biology_description?

 

Always

Should be consistent with platforms used; single, multiple hybridisations; ordered/ unordered; serial; grouping; replicates.

The element hybridization references labelled_extract (the ultimate derived sample). derived_sample references other derived_samples and primary_sample.

Each sample and each array are assigned an (unique) ID. A matrix of IDs should be provided indicating which sample IDs were used on which array IDs.

Replicates

Hybridisations done using the same samples.

replicate_info

Multiple hybridisations

referenced by the element: quality

text description in replicate_info; experimental_design: quality: has_replicates = “true”

Peer reviewed publication

A publication based on the  experimental study has  been accepted in a peer-reviewed journal.

publication

When one exists

experiment_set references publication and experimental_design: quality

publication: citation or dbxref attribute list

experimental_design: quality: peer_reviewed = “true”

Number of replicates

The number (or range) of times each sample has been replicated.

can be put in replicate_info

Replicates were done

Should be consistent with sample and arrays list.

referenced by the element: quality

Number of replicates for each sample or minimum and maximum numbers of sample replicates.

Other quality control steps

Other measures taken to ensure or measure quality.

quality_info

When these have been done

referenced by the element: quality

text description in quality_info

Qualifier, value, source (may use more than once)

Describe any further   information about the experiment set in a structured manner. 

Parameter (for qualifier, value)

Reference (for source)

When additional information is available that would be useful to base queries on.

Should be consistent with other values

Qualifier= parameter:name

Value= parameter:value

Source= reference:description, dbxref (ontology, dictionary, or controlled vocabulary that terms were drawn from).

Experiment set description

Describe any further   information about the experiment set in free text. 

biology_description;

analysis_description

When additional information is available.

Should be consistent with other experiment_set and experiment_design elements and attributes

Free text

 

 

Description

MAML Data format

When applicable

Dependencies

Allowed values

Array design

Description of each array used and each array element (spot).

 

 

 

 

Array design name

Submitter’s name for the array design (e.g., Stanford Human 10K set, lymphochip)

Array_definition

Always

Should not be the same as other submitter’s names unless it is the same array design.

Referenced by array_platform. References array_element_list.

Array_definition attributes: model_name, version

Platform type

Indicator of whether the array is made by spotting or by in situ synthesis

None

Always

Should be consistent with platforms used

In situ synthesized, spotted

Provider

The manufacturer of the array. May be the submitter.

contact

Always.

Should be consistent with platforms used.

Referenced by array_definition.

Contact element attribute list.

Surface type

Example types are glass, membrane.

Array_definition

When array design is novel (i.e., cannot refer to manufacturer).

 

Should be consistent with platform type

Referenced by array_platform. References array_element_list.

Array_definition attribute: surface_type = glass, membrane, other.(replaces absorptive, nonabsorptive)

Surface type name

A descriptive name such as “in-house coated EMBL slides”

Array_definition

When array design is novel (i.e., cannot refer to manufacturer).

Surface type

Referenced by array_platform. References array_element_list.

Array_definition attribute: surface_type_name  = brief text

Array dimensions

The length of each of the array’s two dimensions.

Array_definition

When array design is novel (i.e., cannot refer to manufacturer).

Referenced by array_platform. References array_element_list.

Array_definition attributes: short_axis_length, long_axis_length

Number of array elements

The number of  (composite) elements in the array.

Array_definition

When array design is novel (i.e., cannot refer to manufacturer).

Referenced by array_platform. References array_element_list.

Array_definition attribute: number_of_elements

Reference system

The arrangement and system used to specify the location for each element on the array.

Array_element attributes: block, row, column?

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with array dimensions and number of array elements.

Referenced by array_element_list.

Type of element arrangement (e.g., grid, row, column) and numbering for each type.

Provider ID

A unique ID assigned by the manufacturer such as a serial number.

Array_platform

Always

Should be consistent with samples and arrays list. References array_definition.

Array_platform

Production protocol

A description of how the array was manufactured. Should include origin of reference system (e.g., upper left).

Protocol

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with reference system.

Referenced by array_element_list.

Protocol subelements and attributes.

type = “array_printing”

Hardware subelements and attributes.

Type = “array_printer”

Qualifier, value, source (may use more than once)

Describe any further   information about the array in a structured manner. 

Parameter (for qualifier, value)

Reference (for source)

When additional information is available that would be useful to base queries on.

Should be consistent with other values

Referenced by array_definition.

Qualifier= parameter:name

Value= parameter:value

Source= reference:description, dbxref (ontology, dictionary, or controlled vocabulary that terms were drawn from).

Element position

Location on the array that can be used to identify the element on an image of the array.

Array_element

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with reference system.

Array_element attributes: block, row, column, x_microns, y_microns

Element type

Physical nature of the elements.

Array_element_list

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with platform type.

Array_element_list attribute: derivation = empty, pcr, synthesized_oligo, intact_plasmid, colony, other

Element clone information

For each element, the identity of the clone along with information on the clone provider, the date obtained, and availability.

Clone

When elements are obtained from clones and when array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with element type.

Referenced by array_element.

Clone: db_xref and name attribute (identity); description (provider, date obtained, availability)

Element sequence information

The nucleotide sequence located at each element.

Actual_sequence;

reference sequence

When elements are NOT composite and when array design is novel (i.e., cannot refer to manufacturer) or when additional is available.

Should be consistent with element type and clone. Referenced by array_element, also by clone and gene.

actual_sequence and

reference sequence subelements and attributes.

Specify start and stop if use reference sequence.

Composite element sequence information

The number of distinct sequences in each composite element and the sequences.

array_element: type = multiple_sequence

When elements ARE composite and when array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with element type.

inferred from number of actual_sequence and reference_sequences associated with an array_element.

Single or double stranded

Whether the element sequences are single or double stranded.

Array_element_list

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with element type.

Array_element_list attribute: strandedness = single, double

Element dimensions

The physical dimensions of each element.

Array_element_list

When array design is novel (i.e., cannot refer to manufacturer).

Should be consistent with array dimensions and number of array elements.

Array_element_list attribute: diameter_target (implies circle, why not 2 axes?)

Nucleotide attachment

How the element sequences are physically attached to the array.

Array_element_list<