Uploading files¶

You can head to the Files tab to manage your files. From there, you will be able to upload new files (from your local computer, or a remote endpoint), and delete them.

Data visibility¶

By default, all your uploaded files and datasets are private. If you have administrator privileges, you can select the Integrate (Public dataset) button during integration to make the dataset Public

CSV/TSV files¶

AskOmics will integrate a CSV/TSV file using its header. The type of each column will be predicted, but you will be able to modify it before integration.

Entity (first column)¶

Entity URI¶

The first column of the file will manage the entity itself : the column name will become the entity name, and the values will become the entity's instances URIs.
URIs will be created as follows :

• If the value is an URL, it will be integrated as it is.
• If the value is a CURIE, it will be transformed into an URL before integration. The list of managed CURIE formats is available here.
• Else, the value will be added to either AskOmics namespace_data value, or a custom base URI if specified in the integration form.

Entity type¶

The entity type can either be "starting entity", or "entity". If "starting entity", it may be used to start a query on the AskOmics homepage. Both types will appear as a node in the AskOmics interface.

Inheritance¶

The entity can inherit the attributes and relations of a 'mother' entity. Meaning, you will be able to query the sub-entity on both its own, and its 'mother' attributes and relations. The 'mother' entity however will not have access to any 'daughter' attributes or relations.

To setup inheritance, the column name needs to be formated as follows:
- daughter_entity_name < mother_entity_name (with the < symbol) ie: Custom_population < General population

Entity label (first and second column)¶

To manually set an entity label, you can set the second column as a Label column.
The values of this column will be used as labels for the generated entities.

If there is no Label column, the labels will be generated based on the URIs (The first column).

• If the value is an URL, the last non-empty value after a "/" or "#" will be the label.
• If the value is a CURIE, the value after ":" will be the label
• Else, the raw value is the label

Attributes¶

Each column after the first one will be integrated as an attribute of the entity. The column name will be set as the name of the attribute.
Several attribute types are available. The type of an attribute will dictate the way it will be managed in the query form (eg: text field, value selector...)

!!! note 'Info' AskOmics will try to guess the type of a column based on its name and its values. You will be able to set it manually if the auto-detected type doesn't fit.

Attributes can be of the following types :

Base types¶

• Numeric: if the values are numeric
• Text: if all values are strings
• Date: if all values are dates (using dateutil.parser)
  • Auto-detected terms are 'date', 'time', 'birthday', 'day'
• Category: if there is a limited number of repeated values
• Boolean: if the values are binary ("True" and "False", or "0" and "1")

FALDO types¶

If the entity describe a locatable element on a genome (based on the FALDO ontology):

• Reference: chromosome (Auto-detected terms : 'chr', 'ref', 'scaff')
• Strand: strand (Auto-detected terms : 'strand')
• Start: start position (Auto-detected term : 'start', 'begin')
• End: end position (Auto-detected terms : 'end', 'stop')

Relations¶

A column can also symbolize a relation to another entity. In this case, the column name must be of the form :

• relationName@RelatedEntityName (with the @ symbol)
  • ie: Derives_from@Gene

Two types are available :

• Directed: Relation from this entity to the targeted one (e.g. A is B’s father, but B is not A’s father)
• Symetric: Relation that works in both directions (e.g. A loves B, and B loves A)

Linked URIs must match one of these three formats :

• Full URI
• CURIE
• Simple value (the value will transformed into an URI with AskOmics namespace_data value)

This link between entities will show up in the query screen, allowing users to query related entities.

GFF files¶

Each GFF file can be integrated into several entities. You will be able to select the entities you wish to integrate beforehand. Available entities are the values of the 'type' column of the GFF file. The relations between entities (eg: Parents or Derives_from) will also be integrated.

Extracted attributes are the following :

• Reference
• Strand
• Start
• End
• Any attribute in the attributes column
  • Parents and Derives_from will be converted in relations

BED files¶

Each BED file will be integrated into one entity (the default entity name will be the file name, but it can be customized).

Extracted attributes are the following :

• Reference
• Strand
• Start
• End
• Score

TTL Files¶

You can integrate TTL files in AskOmics, either to integrate your own data, or to enable federated queries to remote endpoints.
In both case, you will need to generate or convert your data in AskOmics's format.

This can be done either manually or automatically