Integration

FireBrowse integration

FireBrowse is the Broad Institute GDAC's public API over The Cancer Genome Atlas (TCGA) data processed by its Firehose pipeline, whose latest run is dated 2016-01-28. Read sample-level mRNASeq and miRSeq expression, clinical data elements, MutSig mutations (MAF) and GISTIC2 gene copy number values and calls; list MutSig significantly mutated genes and GISTIC2 significantly amplified or deleted genes; compute a gene's expression quartiles per cohort; find analysis report links and standard data archive download URLs; and translate TCGA codes (cohorts, sample types, barcodes, centers, platforms, tissue source sites). The experimental feature-table route (TSV only) and the server health check are not exposed. No credential is needed.

Connect FireBrowse How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in FireBrowse.Built from the 28 FireBrowse tools below, on the connection you add once.

What you can do with FireBrowse

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Sample type, 4 operations

get_sample_type_by_barcodeGet sample type by barcodeGETReturns the short letter sample type code (TP, TM, NT...) of a TCGA sample barcode. The barcode must include the sample part (TCGA-GF-A4EO-06); a participant barcode alone returns an empty answer.
get_sample_type_by_codeGet sample type by codeGETTranslates numeric TCGA sample type codes (01, 06) into their short letter codes (TP, TM).
get_sample_type_by_letter_codeGet sample type by letter codeGETTranslates short letter TCGA sample type codes (TP, TM) into their numeric codes (01, 06).
list_sample_typesList sample typesGETLists every TCGA sample type with its numeric code, short letter code and definition, for example 01, TP, Primary Solid Tumor.

Gene copy number, 3 operations

get_gene_copy_numberGet gene copy numberGETReturns gene-level GISTIC2 copy number values for samples (all_data_by_genes), in units of copy number minus 2: 0 is no change, positive an amplification, negative a deletion. At least one gene or participant barcode is required.
get_gene_copy_number_callsGet gene copy number callsGETReturns GISTIC2 discrete amplification and deletion calls per gene and sample (all_thresholded_by_genes): 0 none, 1 or 2 amplification, -1 or -2 deletion, 2 and -2 marking events beyond the arm level. At least one gene or barcode is required.
get_gene_focal_copy_numberGet gene focal copy numberGETReturns gene-level GISTIC2 copy number values for samples counting only focal events (focal_data_by_genes), in units of copy number minus 2. At least one gene or participant barcode is required.

Clinical data, 2 operations

get_clinical_dataGet clinical dataGETReturns TCGA patient clinical data elements (CDEs) verbatim, every CDE in its original form, one record per patient. Select by cohort, participant barcode or CDE name; at least one of the three is required.
get_normalized_clinical_dataGet normalized clinical dataGETReturns the patient-level clinical data elements (CDEs) that Firehose normalized for use in its analyses, one record per patient. Select by cohort, participant barcode or CDE name; at least one of the three is required.

Clinical data element, 2 operations

list_clinical_data_elementsList clinical data elementsGETLists the names of all patient-level clinical data elements (CDEs) in TCGA, across all cohorts, usable as cde_name in get_clinical_data. A CDE is listed only when at least one patient has a value other than NA for it.
list_normalized_clinical_data_elementsList normalized clinical data elementsGETLists the names of the patient-level clinical data elements (CDEs) Firehose normalized for its analyses, across all cohorts, usable as fh_cde_name in get_normalized_clinical_data.

Amplified gene, 1 operation

list_amplified_genesList amplified genesGETLists genes in the regions GISTIC2 found significantly amplified (amp_genes.conf_99), with each region's cytoband, q-value, residual q-value and wide peak boundaries. At least one gene or cohort is required.

Analysis report, 1 operation

list_analysis_reportsList analysis reportsGETLists links to the summary reports of Broad Institute GDAC Firehose analysis runs, filtered by cohort, report type and report name. The reports of the latest run (2016_01_28) are returned unless date selects other runs.

Center, 1 operation

list_centersList centersGETLists the TCGA consortium member centers with their domain name, full organizational name, short name, code and type. Pass center to look up only some domains.

Cohort, 1 operation

list_cohortsList cohortsGETLists the TCGA disease cohort abbreviations with their full disease names, for example BRCA for Breast invasive carcinoma. Pass cohort to translate only some abbreviations.

Deleted gene, 1 operation

list_deleted_genesList deleted genesGETLists genes in the regions GISTIC2 found significantly deleted (del_genes.conf_99), with each region's cytoband, q-value, residual q-value and wide peak boundaries. At least one gene or cohort is required.

Maf column, 1 operation

list_maf_columnsList maf columnsGETLists the names of all columns in the mutation annotation files (MAFs) served by FireBrowse, usable in the column parameter of get_mutations.

Mirseq expression, 1 operation

get_mirseq_expressionGet mirseq expressionGETReturns sample-level log2 miRSeq expression values for one or more miRs across TCGA samples. Filter by cohort, participant barcode, sample type or Firehose preprocessing tool. At least one miR is required.

Mrnaseq expression, 1 operation

get_mrnaseq_expressionGet mrnaseq expressionGETReturns sample-level log2 mRNASeq expression values, with z-scores, for one or more genes across TCGA samples. Filter by cohort, participant barcode, sample type or protocol (RSEM or RPKM). At least one gene is required.

Mrnaseq quartiles, 1 operation

get_mrnaseq_quartilesGet mrnaseq quartilesGETComputes the mRNASeq expression quartiles, median, whisker bounds and outliers of one gene for each cohort, ready to draw a boxplot (Tukey 1977). Filter by cohort, sample type or protocol, or exclude participants.

Mutation, 1 operation

get_mutationsGet mutationsGETReturns mutations from the final analysis MAF (mutation annotation file) generated by MutSig. Filter by gene, cohort, MutSig version or participant barcode; at least one gene, barcode or cohort is required.

Patient, 1 operation

list_patientsList patientsGETLists the TCGA patient barcodes in FireBrowse with their cohort, optionally only for some cohorts. The barcodes are read from the clinical data. Results are paginated with page and page_size.

Platform, 1 operation

list_platformsList platformsGETLists the technology platforms used to sequence or characterize TCGA samples, with their short code and full name. Pass platform to translate only some codes.

Run date, 1 operation

list_run_datesList run datesGETLists the date stamps (YYYY_MM_DD) of the GDAC Firehose runs ingested into FireBrowse, latest first. They are the values accepted by the date filter of list_standard_data_archives and get_sample_counts.

Sample count, 1 operation

get_sample_countsGet sample countsGETReturns aliquot counts per cohort, sample type and data type (clinical, copy number, mRNASeq, miRSeq, methylation, RPPA, mutations and more). The sample type Tumor adds up every tumor aliquot into one number per cohort and data type.

Significantly mutated gene, 1 operation

list_significantly_mutated_genesList significantly mutated genesGETLists significantly mutated genes as scored by MutSig, with mutation counts by type, patients mutated and p- and q-values. Filter by cohort, gene, rank, MutSig version or q-value threshold; at least one cohort or gene is required.

Standard data archive, 1 operation

list_standard_data_archivesList standard data archivesGETLists the download URLs and sizes of the Firehose standard data archives, filtered by run date, cohort, data type, tool, platform, center, data level and protocol. The latest run (2016_01_28) is the default.

Tissue source site, 1 operation

list_tissue_source_sitesList tissue source sitesGETLists the tissue source sites (TSS) that contributed tissue to TCGA, with their code, institution, disease studied and Biospecimen Core Resource. Pass tss_code to look up only some sites.

Use FireBrowse from Claude Code, Cursor or Codex

Connect FireBrowse in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its FireBrowse operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect FireBrowse

Create your workspace
Sign up free. Nothing to install.
Connect FireBrowse once
FireBrowse needs no credential, so its tools work immediately.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.