EMX format
Last updated
Last updated
The default import format for MOLGENIS is 'EMX'. This is a flexible spreadsheet format (Excel, CSV) that allows you to annotate your data with a data model. This works because you can tell MOLGENIS the 'model' of your data via a special sheet named 'attributes'. Optionally, you can also add metadata on entities (i.e., classes, tables), and packages (i.e, models and submodels). It is also possible to provide packages in an emx file via the 'packages' sheet without providing the attributes sheet.
(download)
Note: In order to upload data, at least one group must be created. View the section on groups and roles, for information on creating a group.
For example, if you want to upload an Excel with sheet 'patients':
displayName | firstName | lastName | birthdate | birthplace |
---|---|---|---|---|
Then you must provide a model of your 'patients' via Excel with sheet named 'attributes':
name | entity | dataType | idAttribute | nillable | description |
---|---|---|---|---|---|
'entity' should show the name of your data sheet. Each attribute the column headers in your data. Default dataType is 'string' so you only need to provide non-string values (int, date, decimal, etc). And you must always provide one idAttribute that has 'nillable' = 'FALSE'.
You can first upload the 'model' and then the 'data'. Or you can put the both into one file and upload in one go. What you prefer :-) [todo: provide example files for download]
(download)
Lets assume we want to upload multiple data sheets, with relations between them:
Cities:
Patients:
Notes: birthplace refers to elements in the cityName values in the cities table. children contains comma separated values referring to another patient via displayName (trailing spaces will be removed). Warning: when using excel, be sure your decimal separator is a ".", instead of ",", otherwise mrefs might be seen as decimals when their id is a number, this causes molgenis to see a dot between your references and the importer to fail when uploading.
Users:
Note: users looks similar patients, i.e. they are also persons having 'displayName', 'firstName', and 'lastName'. We will use this in the model below.
To model the data advanced data example, again you need to provide the 'attributes' (i.e., columns, properties). Optionally, you can also describe entities (i.e., classes, tables), and packages (i.e, models and submodels) which gives you some advanced options.
Attributes:
The example below defines the model for entities 'city', 'patient' and 'user'. Note that 'users' had some attributes shared with 'patients' so we will use 'object orientation' to say that both 'user' and 'patient' are both a special kind of 'persons'. This will be defined using the 'extends' relation defined in the 'entities' sheet below.
Entities:
In most cases the 'attributes' sheet is all you need. However, in some cases you may want to add more details on the 'entity'. Here we wanted to show use of 'abstract' (i.e., interfaces) to create model class 'persons' and 'extends' (i.e., subclass, inheritance) to define that 'user' and 'patient' have the same attributes as 'persons'. When data model become larger, or when many data sheets are loaded then the 'package' construct enables you to group your (meta)data.
Packages:
For names in the EMX format, the following rules apply:
Name cannot be empty.
Only letters (a-z
, A-Z
), digits (0-9
), underscores (_
) and dashes (-
) are allowed.
The keywords: login
, logout
, csv
, base
, exist
, meta
and _idValue
are not allowed as entity and attribute names.
attribute names also allow the hash character (#
), e.g. #CHROM
is a valid attribute name.
In attribute names, the dash (-
) is reserved for localization, e.g. description-nl
contains the Dutch translation of the description
attribute.
These restrictions only apply to the technical names, labels are not limited by these rules.
Creating a package without a parent package (also known as a root package) automatically results in the creation of a group. Initially the group name is set to the package identifier, unless the package identifier is not a valid group name. In this case a unique group name is generated. The initial group label is set to the package label. Otherwise the group creation is the same as when created using the security manager plugin. Both name and label can be modified afterwards.
View the section on groups and roles, for information on creating a group.
Name of the entity this attribute is part of
Name of attribute, unique per entity.
Defines the data type (default: string)
Used in combination with xref, mref, categorical, categorical_mref or one_to_many. Should refer to an entity.
Whether the column may be left empty. Default: true Can also contain an expression to dynamically decide if the attribute may be left empty or not. See the Expressions section for a syntax description. If visible is an expression that does not evaluate to true, the attribute may be left empty.
Whether this field is the unique key for the entity. Default: false. Use 'AUTO' for auto generated (string) identifiers.
Whether the value for this field is automatically generated. Default: false. Can be set to true when idAttribute is true or data type is one of [string, data, datetime].
Free text documentation describing the attribute
Description for specified language (can be multiple languages, example: description-nl)
The maximum number of characters.
For string
, email
, enum
and hyperlink
, the default is 255.
For text
and script
, the default is 64 KiB.
Used to set range in case of int or long attributes
Used to set range in case of int or long attributes
true/false, default false
Indicates if this attribute should appear in the xref/mref search dropdown in the dataexplorer. A lookupAttribute must be visible. An entity inherits the lookupAttributes from the entity it extends.
If an entity has no lookupAttributes, the labelAttribute is used in the dropdown.
optional human readable name of the attribute
label for specified language (can be multiple languages, example: label-nl)
true/false to indicate if the user can use this atrribute in an aggregate query
true/false to indicate that the value of this attribute should be used as label for the entity (in the dataexplorer when used in xref/mref). Default: false. A labelAttribute must be visible. If an entity's idAttribute is not visible, it should have a labelAttribute.
true/false to indicate a readOnly attribute
ability to tag the data referring to the tags sections, described below
Validation expression that must return a bool. Must return true if valid and false if invalid. See the Expressions section for a syntax description.
true/false to indicate whether the attribute can be seen by users. Can also contain an expression to dynamically decide if the attribute should be shown or not. See the Expressions section for a syntax description.
value that will be filled in in the forms when a new entity instance is created. For mref and categorical_mref, this should be a comma separated list of ids. For categorical and xref this should be the id of the refEntity. For bool should be true or false. For datetime should be a string in the format YYYY-MM-DDTHH:mm:ssZZ. For date should be a string in the format YYYY-MM-DD.
is used to group attributes into a compound attribute. Put here the name of the compound attribute.
is used to create computed attributes.
Computed object example: "computed myXref" (config attributes table)
Create a two new target attributes (attr1, attr2) in a new entity (newEntity).
Create a xref attribute (myXref) to contain the computed entity.
Add in the expression column of new xref attribute (myXref) the next script: "{attr1: myAttr1, attr2: myAttr2}"
The name of the attributes to convert from should be in the same entity as the new xref attribute (myEntity).
Template
In addition to basic 'computed strings' and 'computed' objects a template can be used as expression. The template expression format is: {"template":"..."} with the value a Mustache template. Tags must refer to attribute identifiers (e.g.
{{myStringAttribute}}). For attributes referencing another entity type the attribute in the referencing entity type needs to be specified as well (e.g.{{myXrefAttribute.id}}).
Example:
Math helper
Use a math helper to do some simple arithmetics. {"template":"{{molgenis-math myIntValue '+' mySecondIntValue}}"}
Will add the 2 attributes.
It's possible to nest multiple math helpers like this: {"template":"{{molgenis-math myIntValue '+' (molgenis-math mySecondIntValue '-' myThirdDecimalValue)}}"}
Valid operators are: '+'
, '-'
, '*'
, '/'
and the modulo operator '%'
unique name of the entity. If packages are provided, name must be unique within a package.
reference to another entity that is extended
name of the group this entity is part of
indicate if data can be provided for this entity (abstract entities are only used for data modeling purposes but cannot accept data)
free text description of the entity
description for specified language (can be multiple languages, example: description-nl)
the backend (database) to store the entities in (currently only PostgreSQL)
ability to tag the data referring to the tags sections, described below
unique name of the package. If parent package is provided the name is unique within the parent.
free text description of the package
use when packages is a sub-package of another package
mechanism to add flexible meta data such as ontology references, hyperlinks
Optionally, additional information can be provided beyond the standard meta data described above. Therefore all meta-data elements can be tagged in simple or advanced ways (equivalent to using RDF triples). For example, above in the packages example there is a 'homepage' tag provided. For example:
unique name of this tag, such that it can be referenced
the human readable label of the tag (e.g. the 'like' tag as shown above).
url to the value object (will become an hyperlink in the user interface)
human readible label of the relation, e.g. 'Documentation and Help'
url to the relation definition, e.g. http://edamontology.org/topic_3061
name of the code system used, e.g. EDAM
You can internationalize attribute labels and descriptions, entity labels and descriptions and you can define internationalized versions of entity attributes.
description-{languageCode} : description for specified language (can be multiple languages) label-{languageCode} : label for specified language (can be multiple languages)
Example:
description-{languageCode} : description for specified language (can be multiple languages) label-{languageCode} : label for specified language (can be multiple languages)
Example:
You can internationalize attributes by postfixing the name with -{countryCode}.
If this is the label attribute, you must set all city-xx labelAttribute values to 'TRUE' on the 'entities' tab.
Example:
entities:
gender:
cityName | lat | lng |
---|---|---|
displayName | firstName | lastName | birthdate | birthplace | children | disease |
---|---|---|---|---|---|---|
userName | active | displayName | firstName | lastName |
---|---|---|---|---|
name | entity | dataType | nillable | refEntity | idAttribute | description |
---|---|---|---|---|---|---|
name | package | extends | abstract | description |
---|---|---|---|---|
name | description | parent |
---|---|---|
Data type | Description | Expected formatting |
---|---|---|
name | entity | label | dataType | idAttribute | refEntity | nillable | visible | labelAttribute | expression |
---|---|---|---|---|---|---|---|---|---|
name | entity | dataType | refEntity | expression |
---|---|---|---|---|
identifier | label | objectIRI | relationLabel | codeSystem | relationIRI |
---|---|---|---|---|---|
name | package | description-en | description-nl | label-en | label-nl |
---|---|---|---|---|---|
name | entity | idAttribute | description-en | description-nl | label-en | label-nl |
---|---|---|---|---|---|---|
name | entity | idAttribute | label-nl | label-de | labelAttribute |
---|---|---|---|---|---|
name | genderlabel-nl | genderlabel-de |
---|---|---|
john_doe
john
doe
1976-03-13
new york
jane_doe
jane
doe
metropolis
papa_doe
papa
doe
new york
displayName
patients
TRUE
FALSE
name
firstName
patients
FALSE
first name
lastName
patients
FALSE
family name
birthdate
patients
date
FALSE
day of birth
birthplace
patients
FALSE
place of birth
new_york
40,712784
-74,005941
metropolis
37,151165
-88,731998
john_doe
john
doe
1976-03-13
new_york
none
jane_doe
jane
doe
metropolis
none
papa_doe
papa
doe
new_york
john_doe, jane_doe
cardio
jdoe
TRUE
john_doe
john
doe
jdoe2
jane_doe
jane
doe
pdoe
papa_doe
papa
doe
cityName
cities
FALSE
TRUE
unique city name
lat
cities
decimal
latitude in degrees
lng
cities
decimal
longitude in degrees
displayName
patients
FALSE
TRUE
unique name
firstName
persons
first name
lastName
persons
family name
children
patients
mref
patients
children of a patient
birthdate
patients
date
day of birth
birthplace
patients
xref
cities
place of birth
disease
patients
disease description
userName
users
FALSE
TRUE
unique login name
active
users
bool
whether user is active
cities
hospital
list of cities
persons
hospital
true
person defines general attributes like firstName, lastName
users
hospital
persons
users extends persons, meaning it 'inherits' attribute definition
patients
hospital
persons
patient extends person, adding patientNumber
root
my main package
hospital
sub package holding entities to describe all kinds of persons
root
string
The default data type in MOLGENIS, a character string.
A string of characters
text
A long character string, when editing the user will be presented with a textarea.
A string of characters
int
Integers. Natural numbers like 1, 2, 3, -1, -2, -3. rangeMin and rangeMax can be defined.
Non decimal numbers in range[-2^31 , 2^31 -1]
long
Non-decimal number of type long
Non decimal numbers in range[-2^63 , 2^63 -1]
decimal
Decimal numbers/floats.
Decimal numbers with double-precision ('NaN' not allowed)
bool
A boolean value: true/false
TRUE/FALSE
date
A date without a time-zone in the ISO-8601 calendar system
yyyy-mm-dd
datetime
An instant in time. Time zone information may be used to specify the instant but is not stored.
yyyy-mm-ddThh:mm:ss+timezone e.g. 1985-08-12T11:12:13+0500
xref
Reference to an attribute of another entity. Using this type requires another entity (refEntity) with information that is linked to the selected entity. Although you should always refer to the id of an attribute defined as xref, in your data explorer the label of the refEntity will be presented instead. When label is not specified, id will be used as label. When searching for a specific xref value in the filter wizard, the value has to be typed partly and then selected out of a list with suggestions.
Id of the attribute you wish to link, this id should always be available in the refEntity.
mref
Reference to several attributes of another entity. Using this type requires another entity (refEntity) with information that is linked to the selected entity. Although you should always refer to the id of an attribute defined as mref, in your visualisation, the defined label for the refEntity will be presented, when label is not given, id will be used as label. When searching for a specific mref value in the filter wizard, the value has to be typed partly and then selected out of a list with suggestions.
A comma separated list of id’s, these id’s should always be available in the refEntity.
categorical
Reference to an attribute of another entity. Using this type requires another entity (refEntity) with information that is linked to the selected entity. Although you should always refer to the id of an attribute defined as categorical, in your visualisation, the defined label for the refEntity will be presented, when label is not given, id will be used as label. This type is typically used when answers are fixed like: “Yes”, “No”, “Unknown”. When searching for a specific categorical value in the filter wizard of molgenis a checkbox can be marked.
Id of the attribute you wish to link, this id should always be available in the refEntity.
categorical_mref
Reference to several attributes of another entity. Using this type requires another entity (refEntity) with information that is linked to the selected entity. Although you should always refer to the id of an attribute defined as categorical_mref, in your visualisation, the defined label for the refEntity will be presented, when label is not given, id will be used as label. When searching for a specific categorical value in the filter wizard of molgenis a checkbox can be marked.
A comma separated list of id’, these id’s should always be available in the refEntity.
compound
This type can be used to group parts of your data together. Your dataset will consist of several compounds all containing certain attributes. Don’t forget to specify the partOfAttribute column for the attributes you wish to put in the compound.
Nothing
file
A file. Create a column of the 'file' data type requires refEntity FileMeta.
Id of the FileMeta entity you wish to link, this id should always be available in FileMeta.
An e-mail adress
E-mail adress
enum
An item chosen from a fixed list of options that can be selected for this data type. The options should be given in an extra column called “enumOptions”. These options cannot be updated without changing meta-data (so deleting the data and meta data and a new upload are required in MOLGENIS 1.x).
A value chosen from the enumOptions list specified as a comma separated list of options in the model.
hyperlink
A link to a website
A link to a website
script
A piece of code, typically edited in a text editor
A string of characters
one_to_many
This data type is only supported in MOLGENIS 2.0. A data type that defines the one to many relationship between two columns in two separate tables. Having this data types requires having another table with an xref column which is linked to the one_to_many. The one_to_many requires a refEntity like the other referring data types, but unlike the others, it also requires a “mappedBy” column. In this column the name of the xref in the other column should be specified. For instance, an author can write several books. Books are stored in one table (called “books”) with “author” as xref and as refEntity “authors”. In the authors table authors are stored with books as one_to_many. The books attribute has the refEntity “books” and is mappedBy “author”.
A comma separated list of id’s. Requires having an xref to this attribute in another table. Since one_to_many attributes are mapped by a xref attribute they cannot be specified in data sheets.
id
myEntity
Id
int
TRUE
FALSE
FALSE
FALSE
myXref
myEntity
New Entity
xref
FALSE
newEntity
TRUE
TRUE
FALSE
{attr1: myAttr1, attr2: myAttr2}
myAttr1
myEntity
My Attr 1
date
FALSE
TRUE
FALSE
TRUE
myAttr2
myEntity
My Attr 2
int
FALSE
TRUE
FALSE
TRUE
attr1
newEntity
Attr 1
string
FALSE
TRUE
FALSE
TRUE
attr2
newEntity
Attr 2
string
TRUE
TRUE
FALSE
TRUE
id
MyEntityType
string
xref0
MyEntityType
xref
MyReference
xref1
MyEntityType
xref
MyReference
computedXref
MyEntityType
string
{"template":"lorum {{xref0.id}} ipsum {{xref1.label}}"}
mref0
MyEntityType
mref
MyReference
mref1
MyEntityType
mref
MyReference
computedMref
MyEntityType
string
{"template":"my {{mref0.id}} text {{mref1.label}}"}
string0
MyEntityType
string
string1
MyEntityType
string
computedString
MyEntityType
string
{"template":"{{string0}} and {{string1}}"}
id
MyReference
string
label
MyReference
string
like
like
homepage
http://www.molgenis.org
http://www.molgenis.org
homepage
docs
http://some.url
http://www.molgenis.org
Documentation and Help
EDAM
http://edamontology.org/topic_3061
cities
hospital
list of cities
lijst van steden
Cities
Steden
persons
hospital
list of persons
lijst van personen
Persons
Personen
displayName
patients
TRUE
Patient name
Naam van de patient
name
naam
firstName
patients
Patient first name
Voornaam van de patient
first name
voornaam
lastName
patients
Patient family name
Achternaam van de patient
family name
achternaam
name
gender
TRUE
genderlabel-nl
gender
Label (nl)
Etikette (nl)
TRUE
genderlabel-de
gender
Label (de)
Etikette (de)
TRUE
Male
Man
Man
Female
Vrouw
Frau
Unknown
Onbekend
Unbekannt