Expression Atlas Tools#
Configuration File: expression_atlas_tools.json
Tool Type: Local
Tools Count: 4
This page contains all tools defined in the expression_atlas_tools.json configuration file.
Available Tools#
ExpressionAtlas_get_baseline (Type: ExpressionAtlasTool)#
List baseline gene expression experiments in EBI Expression Atlas for a given species. The gene…
ExpressionAtlas_get_baseline tool specification
Tool Information:
Name:
ExpressionAtlas_get_baselineType:
ExpressionAtlasToolDescription: List baseline gene expression experiments in EBI Expression Atlas for a given species. The gene argument is used to label the request and suggest a follow-up call, but does NOT filter the returned experiment list (the underlying API has no reliable way to filter experiments by gene) – use GxA_get_experiment_expression with an experiment_accession from the results and gene_id=<gene> to check whether a specific experiment actually has data for that gene. Complements GTEx and HPA. No API key needed.
Parameters:
gene(string) (required) Gene symbol (e.g., ‘TP53’, ‘WDR7’) or Ensembl ID (e.g., ‘ENSG00000141510’)species(string) (optional) Species name (default: ‘homo sapiens’). Also supports ‘mus musculus’, ‘rattus norvegicus’, etc.
Example Usage:
query = {
"name": "ExpressionAtlas_get_baseline",
"arguments": {
"gene": "example_value"
}
}
result = tu.run(query)
ExpressionAtlas_get_experiment (Type: ExpressionAtlasTool)#
Get metadata for a specific Expression Atlas experiment by accession (e.g., E-MTAB-2836): accessi…
ExpressionAtlas_get_experiment tool specification
Tool Information:
Name:
ExpressionAtlas_get_experimentType:
ExpressionAtlasToolDescription: Get metadata for a specific Expression Atlas experiment by accession (e.g., E-MTAB-2836): accession, type, species, and description/URLs. Caveat: the underlying GXA API does not expose per-experiment design/technology/contrasts/PubMed/assay-count data through this endpoint (confirmed live: those fields are always empty) – for that level of detail, visit the experiment’s page on https://www.ebi.ac.uk/gxa directly.
Parameters:
accession(string) (required) Expression Atlas experiment accession (e.g., ‘E-MTAB-2836’, ‘E-GEOD-26284’)
Example Usage:
query = {
"name": "ExpressionAtlas_get_experiment",
"arguments": {
"accession": "example_value"
}
}
result = tu.run(query)
ExpressionAtlas_search_differential (Type: ExpressionAtlasTool)#
Search for differential expression EXPERIMENTS from EBI Expression Atlas by condition text and/or…
ExpressionAtlas_search_differential tool specification
Tool Information:
Name:
ExpressionAtlas_search_differentialType:
ExpressionAtlasToolDescription: Search for differential expression EXPERIMENTS from EBI Expression Atlas by condition text and/or a gene mentioned in the experiment description. Supports disease-context expression analysis. Caveat: condition does real text filtering, but gene is a text-search coincidence (experiment descriptions rarely name individual genes) and does not return actual per-gene differential-expression values – treat gene-tagged results as a starting point for manual review, not a validated gene-specific result set.
Parameters:
gene(string) (optional) Gene symbol or Ensembl ID (e.g., ‘TP53’, ‘ENSG00000141510’)condition(string) (optional) Condition/disease to filter by (e.g., ‘cancer’, ‘inflammation’, ‘breast’)species(string) (optional) Species name (default: ‘homo sapiens’)
Example Usage:
query = {
"name": "ExpressionAtlas_search_differential",
"arguments": {
}
}
result = tu.run(query)
ExpressionAtlas_search_experiments (Type: ExpressionAtlasTool)#
Search all EBI Expression Atlas experiments (baseline + differential) by condition text and/or a …
ExpressionAtlas_search_experiments tool specification
Tool Information:
Name:
ExpressionAtlas_search_experimentsType:
ExpressionAtlasToolDescription: Search all EBI Expression Atlas experiments (baseline + differential) by condition text and/or a gene mentioned in the experiment description. Returns experiment accessions, types, descriptions, and assay counts. Caveat: condition does real text filtering, but gene is a text-search coincidence (experiment descriptions rarely name individual genes), so a gene-only query effectively returns the unfiltered catalog – use to find relevant expression datasets, not as a validated per-gene result set.
Parameters:
gene(string) (optional) Gene symbol or Ensembl IDcondition(string) (optional) Biological condition or tissue (e.g., ‘brain’, ‘cancer’, ‘kidney’)species(string) (optional) Species name (optional, e.g., ‘homo sapiens’)
Example Usage:
query = {
"name": "ExpressionAtlas_search_experiments",
"arguments": {
}
}
result = tu.run(query)