Gxa Tools#

Configuration File: gxa_tools.json Tool Type: Local Tools Count: 3

This page contains all tools defined in the gxa_tools.json configuration file.

Available Tools#

GxA_get_experiment_expression (Type: GxATool)#

Get gene expression data from a specific Expression Atlas experiment. Returns expression levels a…

GxA_get_experiment_expression tool specification

Tool Information:

  • Name: GxA_get_experiment_expression

  • Type: GxATool

  • Description: Get gene expression data from a specific Expression Atlas experiment. Returns expression levels across tissues or conditions. IMPORTANT limitation (verified live): the upstream /experiments/{accession} endpoint always returns only a small, arbitrary default SAMPLE of the experiment’s genes (e.g. ~29 of 9,570 for E-MTAB-2836) and silently ignores every gene-filter parameter we can send it – so the optional gene_id filter can only ever match within that tiny sample, never the full gene set. The response discloses profiles_returned_by_upstream and profiles_available_upstream so callers can see the sample size vs. the true total; when a gene_id filter matches nothing, coverage_warning explains that this is NOT evidence the gene is unexpressed – it may simply have fallen outside the sample. For a reliable per-gene answer, use GTEx_get_expression_summary (human tissues) or HPA_get_rna_expression_in_specific_tissues instead. Experiment E-MTAB-2836 contains RNA-seq of 122 human tissues, E-MTAB-2706 covers 675 cancer cell lines.

Parameters:

  • experiment_accession (string) (required) Expression Atlas experiment accession (e.g., ‘E-MTAB-2836’ for human tissues RNA-seq, ‘E-MTAB-2706’ for cancer cell lines).

  • gene_id (string) (optional) Optional Ensembl gene ID to look for within the small default SAMPLE of genes the upstream API returns for this experiment (e.g. ‘ENSG00000141510’ for TP53). This is NOT a true server-side or full-experiment filter – the upstream endpoint ignores gene-filter parameters and only ever returns a default sample (commonly ~20-30 genes) of the experiment’s full gene set (which can be thousands). A miss does not mean the gene is absent or unexpressed; check coverage_warning in the response. Leave empty to see the same default sample unfiltered.

Example Usage:

query = {
    "name": "GxA_get_experiment_expression",
    "arguments": {
        "experiment_accession": "example_value"
    }
}
result = tu.run(query)

GxA_get_experiment_info (Type: GxATool)#

Get metadata about a Gene Expression Atlas experiment including species, number of genes and cond…

GxA_get_experiment_info tool specification

Tool Information:

  • Name: GxA_get_experiment_info

  • Type: GxATool

  • Description: Get metadata about a Gene Expression Atlas experiment including species, number of genes and conditions, and experimental design. Useful for understanding what data is available before querying specific gene expression. Example: E-MTAB-2836 has 29 conditions (tissues) and profiles for thousands of genes in Homo sapiens.

Parameters:

  • experiment_accession (string) (required) Expression Atlas experiment accession (e.g., ‘E-MTAB-2836’, ‘E-MTAB-2706’, ‘E-GEOD-46817’).

Example Usage:

query = {
    "name": "GxA_get_experiment_info",
    "arguments": {
        "experiment_accession": "example_value"
    }
}
result = tu.run(query)

GxA_list_experiments (Type: GxATool)#

List gene expression experiments from EBI Gene Expression Atlas. Can filter by species (e.g., ‘Ho…

GxA_list_experiments tool specification

Tool Information:

  • Name: GxA_list_experiments

  • Type: GxATool

  • Description: List gene expression experiments from EBI Gene Expression Atlas. Can filter by species (e.g., ‘Homo sapiens’) and experiment type (‘baseline’ for tissue expression, ‘differential’ for condition comparisons). The Atlas contains 4,500+ experiments across hundreds of species. Example: filtering by ‘Homo sapiens’ and ‘baseline’ returns 123 human tissue-level expression experiments including E-MTAB-2836 (122 human tissues), E-MTAB-2706 (675 cancer cell lines), and E-GEOD-46817 (melanoma).

Parameters:

  • species (string) (optional) Species name to filter experiments (e.g., ‘Homo sapiens’, ‘Mus musculus’, ‘Arabidopsis thaliana’). Leave empty for all species.

  • experiment_type (string) (optional) Filter by experiment type: ‘baseline’ (tissue/cell type expression), ‘differential’ (condition comparisons), ‘proteomics’, or ‘rnaseq’. Leave empty for all types.

  • limit (integer) (optional) Maximum number of experiments to return (1-100). Default: 20.

Example Usage:

query = {
    "name": "GxA_list_experiments",
    "arguments": {
    }
}
result = tu.run(query)