senotype-library

Senotype Library

User Documentation

• Senotype Library
  • User Documentation
• Introduction
  • Senotype
  • The SenLib Database
  • The Senotype Library
• User Authentication
• Senotype Library faceted search
• Business Rules of the Senotype Library
  • Edit state
  • Creating a new senotype
• Using the Senotype Definition tools
  • Required data
  • Submission information: name and description
  • Assertions
  • Valueset-based assertions
  • Context assertions
  • External assertions
    • Edit page
      • Translation of codes
      • Details button
    • Search window
      • Search box
      • Explore button
  • Details for specific external assertions
    • SenNet datasets
    • Origin (SciCrunch RRID)
    • Marker assertions
      • Specified markers
      • Regulating markers
      • Collapsible section
      • Search windows (individual addition)
    • Specified markers
    • Regulating markers
• Validation

Introduction

Senotype

A senotype is a senescent cell subtype with distinct transcriptional, spatial, and functional characteristics. Each senotype is described by a cellular senescence functional definition and associated multidimensional characteristics. A senotype refers to the complex interplay of features that associates a cell phenotype with other characteristics that vary widely by tissue type, cell lineage, inducing stimulus, and senescence hallmarks such as cell cycle arrest, SASP production, DNA damage, and chromatin remodeling.

A senotype associates a phenotype with characteristics that include:

• taxa
• locations in the body
• cell types
• microenvironments
• inducers
• hallmarks
• assays
• citations
• origins (RRIDs)
• SenNet datasets
• a set of markers (genes or proteins)
• contextual data, such as age or BMI

A senotype can be represented as a knowledge graph of assertions–i.e., discrete relationships between the senotype and other entities. For example, a senotype can assert a in_taxon relationship with an entity representing “human”.

The entities and relationships of a senotype can be encoded–i.e., represented with codes from standard biomedical vocabularies.

For example, the gene and protein markers associated with a senotype can be identified with codes from HGNC and UniProt, with a relationship encoded by the Relations Ontology.

The SenLib Database

Senotype definitions are maintained in the Senotype Library (SenLib). Senotype definitions are stored in SenLib in JSON format and conform to the schema defined in the senlib GitHub repository.

The Senotype Library

The Senotype Library application allows users to manage senotype definitions in the SenLib database. Because the majority of the assertions in a senotype definition involve categorical data, most of the work of defining a senotype will involve the selection of values from lists.

With the Senotype Library, users can:

1. review a senotype definition
2. create a new senotype definition
3. revise a senotype definition

User Authentication

Only users with Globus accounts with proper authentication will be able to use the Senotype Library.

Upon startup, the Senotype Library will display a login page. The user will be authenticated in Globus. If the user’s Globus account has the necessary privileges, the user will be able to continue to the Edit page.

Senotype Library faceted search

The Senotype Library faceted search allows users to find senotypes either through free text or faceted search options.

Business Rules of the Senotype Library

A user can only edit a senotype submission file for which they are authorized. Authorization is controlled by means of the SenNet Consortium’s Globus environment.

Edit state

Only authorized users may edit a senotype. If a senotype shows a clickable Edit button, the user is authorized to edit the submission file.

Creating a new senotype

User can click on Register Senotype in the top right corner of the application to define a new senotype. The editor will load into the Definition section defaults for the new submission, including:

1. a new SenNet ID
2. the Globus name and email of the user, as submitter

Using the Senotype Definition tools

Required data

A senotype definition requires a minimal set of data. Data that are required for a senotype are indicated with a red asterisk.

Submission information: name and description

The title and description of the senotype definition are required data.

Both the title and description fields are free text.

Assertions

The Definition section allows the definition of assertions between a senotype and its characteristics.

Each input in the Definition section corresponds to an assertion in the senotype definition. For example, the Taxon input corresponds to the senotype definition’s in_taxon assertion.

There are four types of assertions. Each type has its own tool for selection.

Valueset-based assertions

Most of a senotype definition’s assertions will be categorical, in which the possible values for the object of the assertion will be in a valueset. For example, the categories for a senotype definition’s taxon might be the valueset {human, mouse}.

The elements of valuesets are codes from biomedical vocabularies.

Valueset-based assertions include:

type of assertion	source vocabulary
taxon	NCBI
hallmark	SENOTYPE
microenvironment	SENOTYPE
inducer	SENOTYPE
assay	OBI
sex	SNOMED_CT

Because there can be multiple assertions of the same type, the inputs for valueset-based assertions are lists. For example, a senotype can assert relationships with multiple taxa:

Context assertions

Context assertions are allow the user to define numerical ranges for an assertion.

For example, the age assertion can be bounded to apply only to the range of 18 to 89 years.

The current types of context assertions are

• age
• BMI

External assertions

A number of the objects of assertions are encoded with codes from large vocabularies or ontologies, such as Cell Ontology. Instead of using internal valuesets for these inputs, the Senotype Library obtains codes by searching external sources.

assertion type	source	vocabulary
location (organ)	UBKG	Uberon
cell type	UBKG	Cell Ontology (CL)
diagnosis	UBKG	Disease Ontology (DOID)
citation	NCBI PubMed	PMID
origin (RRID)	SciCrunch	RRID
dataset	SenNet Data Portal	SenNet ID
gene marker	UBKG	HGNC
protein marker	UBKG	UniProtKB

Edit page

Translation of codes

Senotype definitions represent entities with codes. For example, if a senotype asserts a relationship with cell type “fibroblast of lung”, its definition represents the cell type with the code CL:0002553.

When the Edit page loads information for an existing senotype, it translates codes into terms by searching external sources. The terms are only to provide human-readable descriptions: they are not stored in SenLib.

Details button

The button next to an assertion object links to a detail page in the source that corresponds to the object.

For example, the details button next to a cell type opens the corresponding page for the cell type in EMBL-EBI’s Ontology Lookup Service:

assertion object type	reference page
location	SenNet Organs detail
cell type	EMBL-EBI Ontology Lookup Service (OLS)
diagnosis	Disease Ontology
citation	PubMed
origin	RRID Portal (see below)
dataset	SenNet Data Portal
gene marker	HGNC
protein marker	UniprotKB

Search window

Search box

In the Search window for an external assertion object, entering input into the search box triggers a search against an external source. The results of a search appear below the search box as a list of links. Selecting a link from the result list will add the result to the corresponding object list in the Edit page.

The forms of search terms depend on the external source. For example, to search for a cell type, a user can enter either the Cell Ontology ID or a portion of the preferred term (e.g,. “fibroblast”)

input type	search source	forms of search term	example
citation	NCBI EUtils API	exact PMID; portion of publication title	41247924; predictors
origin	RRID Portal (SciCrunch)	exact RRID	4850064
dataset	SenNet entity-api; SenNet Data Portal	exact SenNet ID	SNT699.FVQD.882
location	hs-ontology API (UBKG API)	portion of SenNet organ name	lung
diagnosis	hs-ontology API (UBKG API)	exact DOID; exact match for a preferred term or synonym	DOID:3083; 3083; chronic obstructive pulmonary disease; copd
celltype	hs-ontology API (UBKG API)	exact CLID; portion of a preferred term	CL:4006000; 4006000; fibroblast
gene	hs-ontology API (UBKG API)	exact: HGNC ID; approved symbol; alias; prior approved symbol; or prior alias	HGNC:1100; 1100; BRCA1; BRCC1
protein	hs-ontology API (UBKG API)	exact: UniProtKB ID; UniProtKB entry name	UNPROTKB:Q13201; Q13201; MMRN1_HUMAN

Explore button

In an external search window, the button opens the corresponding home page of an external site to facilitate finding an appropriate identifier. For example, the Explore button for the cell type selection opens the page of the OLS Search for the Cell Ontology.

The Senotype Library is not integrated with these external sites. Once the user finds an appropriate identifier or search term, they can enter the id or term into the search box of the Select window.

Details for specific external assertions

SenNet datasets

The Search feature for datasets will direct to the Senotype Consortium pages, including the Data Portal.

By default, the Senotype Data Portal displays only published datasets. If the user wishes to associate a senotype with a dataset that is not published, they will need to log into the Data Portal again with their SenNet consortium user ID.

Origin (SciCrunch RRID)

A senotype definition can be associated with an origin, or Research Resource Identifier (RRID). RRIDs are managed in the RRID Portal. The RRID Portal uses the SciCrunch API for searches.

RRIDs are organized by resource type:

• Cores, Instruments, and Tools
• Plasmids
• Biosamples
• Cells
• Antibodies
• Organisms

The Search feature of the Origin section directs to the RRID Portal. Once the user identifies a RRID, they must enter the ID in the search window. The ‘RRID:’ part of the ID is optional; however, any other letters are required (e.g., “SCR”, “AddGene”, etc.).

Antibodies and cell lines have identifiers with a higher level of resolution than the RRID because they also identify vendors. The Search feature displays the names provided by vendors to distinguish different offerings of the same origin.

The Details page for an origin will open a link to the SciCrunch detail page for the origin. Depending on the type of origin, the SciCrunch detail page may redirect to another site, such as that of a vendor.

Marker assertions

A senotype definition can have two types of markers. The association of a senotype definition with a marker is a statement of tentative expert opinion. Associations are not just the result of a single experiment, but also not necessarily clear evidence.

Specified markers

These gene or protein markers are a list an investigator might recommend be used to describe the markers that would help identify or characterize a senescent cell of this senotype–e.g., to use as a gene panel for a probed assay.

Regulating markers

These are typically a longer list of gene or protein markers that have been tested for the senotype. The investigator observes these markers to be up-regulated; down-regulated; or tested but inconclusive whether up- or down- regulated (e.g., using log2FC and p-value).

Collapsible section

Because it is anticipated that a senotype will be associated with many markers, the entire Markers section is collapsible.

Marker assertion management for both specified and regulating markers is similar: Clicking on Add marker will open a modal allowing the addition of a single or bulk import of markers.

Search windows (individual addition)

A marker Search window searches for gene or protein markers.

The user can enter different types of identifiers for markers, including:

1. HGNC numbers (e.g., 7178)
2. HGNC approved symbols (e.g., MMRN1)
3. HGNC aliases (ECM)
4. HGNC prior symbols (e.g., MMRN)
5. UniProtKB ID (e.g., Q13201)
6. UniProtKB entry name (e.g., MMRN1_HUMAN)

Note case for gene identifiers:

1. HGNC identifiers combine uppercase letters and numbers
2. MGI numbers use mixed case for letters.

The Search window for regulating markers includes a field for type of regulating:

• upregulation
• downregulation
• inconclusive regulation

Both of these modals support bulk import which allows the user to load a large number of markers from a local CSV file that the user specifies.

The bulk addition windows will only add information from a CSV if:

• the CSV has the expected format
• all markers in the CSV can be found in an external source

Specified markers

The CSV used for bulk upload of specified markers must have the following structure:

column	values
type	either gene or protein
id	* if a gene, the HGNC symbol * if a protein, the UniProtKB ID

Example:

type,id
gene,BRCA1
protein,Q13201

Regulating markers

The CSV used for bulk upload of regulating markers must have the following structure:

column	values
type	either gene or protein
id	* if a gene, the HGNC symbol * if a protein, the UniProtKB ID
action	one of the following: * 1 for up regulation * 0 for inconclusive regulation * -1 for down regulation

Example:

type,id,action
gene,BRCA1,1
protein,Q13201,0
gene,BRAF,-1

Validation

When the user clicks the Update/Submit button at the bottom of page, the application sends the form to the Senotype API for validation.

If the data is invalid or incomplete, the :

• displays an error message in red at the top of the Library page
• displays an error message in red next to the field with an issue

The Library verifies that:

1. All required fields have at least one value.
2. Context assertions are ordered such that lowerbound <= value <= upperbound.

If the data passed validation, the Library:

• writes the data to the SenLib database
• displays a success message in green at the top of the Library page
• resets the Navigator to point to the updated submission