{"slug":"gtex-expression-2","title":"gtex-expression","summary":"Guide for integrating GTEx tissue expression data with ENCODE regulatory elements. Use when users need to check if a gene is expressed in a tissue, correlate regulatory elements with expression, or validate ENCODE findings against GTEx. Trigger on: GTEx, tissue expression, gene e","platform":"Claude","tags":[],"authorName":"LLM Mart","authorSlug":"llm-mart","score":0,"source":"github","price":null,"verified":false,"createdAt":"2026-09-22T13:25:41.991552Z","repo":{"url":"https://github.com/ammawla/encode-toolkit","stars":20,"forks":5,"license":"AGPL-3.0","updatedAt":"2026-09-27T01:22:42Z"},"bodyHtml":"<hr>\n<h2>name: gtex-expression\ndescription: \"Guide for integrating GTEx tissue expression data with ENCODE regulatory elements. Use when users need to check if a gene is expressed in a tissue, correlate regulatory elements with expression, or validate ENCODE findings against GTEx. Trigger on: GTEx, tissue expression, gene expression levels, expression atlas, eQTL, tissue-specific expression, TPM values.\"</h2>\n<h1>Integrating GTEx Tissue Expression with ENCODE Regulatory Data</h1>\n<p>Use GTEx gene expression across 54 human tissues to validate ENCODE regulatory element activity, establish enhancer-gene links, and provide tissue-specific expression context for functional genomics findings.</p>\n<h2>Scientific Rationale</h2>\n<p><strong>The question</strong>: \"Is the gene near my ENCODE regulatory element actually expressed in the tissue where the element is active?\"</p>\n<p>ENCODE catalogs where regulatory elements exist (enhancers, promoters, insulators) but does not directly measure gene expression across a broad tissue panel. GTEx (Genotype-Tissue Expression) fills this gap by providing RNA-seq-based gene expression measurements across 54 human tissues from ~1,000 post-mortem donors. Integrating the two answers a fundamental question: does the regulatory landscape match the transcriptional output?</p>\n<p>An active enhancer (H3K27ac+, ATAC-seq+) near a gene in pancreas tissue is much more meaningful if GTEx confirms the gene is highly expressed in pancreas. Conversely, an ENCODE enhancer near a gene with zero expression in the relevant tissue suggests the enhancer regulates a different gene, or acts in a cell-type subpopulation not captured by bulk GTEx.</p>\n<h3>What GTEx Provides</h3>\n<ul>\n<li><strong>Median gene expression (TPM)</strong> across 54 human tissues from ~1,000 donors</li>\n<li><strong>Transcript-level expression</strong> for isoform analysis</li>\n<li><strong>eQTLs</strong> — variants associated with gene expression in specific tissues (cis and trans)</li>\n<li><strong>sQTLs</strong> — variants associated with alternative splicing</li>\n<li><strong>Single-nucleus RNA-seq</strong> for selected tissues (GTEx v8+)</li>\n<li><strong>Allele-specific expression</strong> data</li>\n</ul>\n<h3>Why Integrate with ENCODE</h3>\n<table>\n<thead>\n<tr>\n<th>ENCODE provides</th>\n<th>GTEx provides</th>\n<th>Together</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Where regulatory elements are</td>\n<td>Where genes are expressed</td>\n<td>Regulatory element-expression correlation</td>\n</tr>\n<tr>\n<td>Tissue-specific enhancers</td>\n<td>Tissue-specific expression</td>\n<td>Enhancer-gene validation</td>\n</tr>\n<tr>\n<td>TF binding sites</td>\n<td>eQTLs in those sites</td>\n<td>Functional variant identification</td>\n</tr>\n<tr>\n<td>Chromatin accessibility</td>\n<td>Expression levels</td>\n<td>Accessibility-expression concordance</td>\n</tr>\n</tbody>\n</table>\n<h2>Key Literature</h2>\n<ul>\n<li><strong>GTEx Consortium 2020</strong> \"The GTEx Consortium atlas of genetic regulatory effects across human tissues\" (Science, ~4,000 citations). The flagship publication describing the v8 release with 17,382 samples across 54 tissues from 948 donors. Identified cis-eQTLs for 95% of genes, tissue-sharing patterns, and cell-type interaction eQTLs. <a href=\"https://doi.org/10.1126/science.aaz1776\">DOI: 10.1126/science.aaz1776</a></li>\n<li><strong>Aguet et al. 2017</strong> \"Genetic effects on gene expression across human tissues\" (Nature, ~3,000 citations). The v6p analysis establishing the multi-tissue eQTL framework, demonstrating widespread tissue-specific genetic regulation. <a href=\"https://doi.org/10.1038/nature24277\">DOI: 10.1038/nature24277</a></li>\n<li><strong>ENCODE Project Consortium 2020</strong> (Nature, ~1,656 citations). Registry of 926,535 human cCREs. Provides the regulatory element catalog to cross-reference with GTEx expression. <a href=\"https://doi.org/10.1038/s41586-020-2493-4\">DOI: 10.1038/s41586-020-2493-4</a></li>\n<li><strong>Nasser et al. 2021</strong> (Nature, ~468 citations). ABC model linking enhancers to genes using ENCODE data. GTEx expression validates ABC-predicted enhancer-gene pairs. <a href=\"https://doi.org/10.1038/s41586-021-03446-x\">DOI: 10.1038/s41586-021-03446-x</a></li>\n</ul>\n<h2>When to Use This Skill</h2>\n<table>\n<thead>\n<tr>\n<th>Scenario</th>\n<th>How GTEx helps</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Found enhancer near gene X in ENCODE</td>\n<td>Check if gene X is expressed in the matching tissue</td>\n</tr>\n<tr>\n<td>GWAS variant in ENCODE peak</td>\n<td>Query GTEx eQTLs to identify regulated gene</td>\n</tr>\n<tr>\n<td>Comparing regulatory landscapes across tissues</td>\n<td>Validate that differential enhancers correspond to differential expression</td>\n</tr>\n<tr>\n<td>Designing functional validation</td>\n<td>Confirm gene is expressed before investing in CRISPR/reporter assays</td>\n</tr>\n<tr>\n<td>Interpreting TF ChIP-seq</td>\n<td>Check if TF target genes show expected expression patterns</td>\n</tr>\n<tr>\n<td>Choosing relevant ENCODE biosamples</td>\n<td>Use GTEx to identify which tissues express your gene of interest</td>\n</tr>\n</tbody>\n</table>\n<h2>GTEx REST API Reference</h2>\n<p><strong>Base URL</strong>: <code>https://gtexportal.org/api/v2</code></p>\n<p>No authentication required. Responses are JSON.</p>\n<h3>Key Endpoints</h3>\n<table>\n<thead>\n<tr>\n<th>Endpoint</th>\n<th>Purpose</th>\n<th>Key Parameters</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td><code>/expression/geneExpression</code></td>\n<td>Median TPM by tissue for a gene</td>\n<td><code>geneId</code>, <code>datasetId</code></td>\n</tr>\n<tr>\n<td><code>/expression/medianTranscriptExpression</code></td>\n<td>Transcript-level TPM by tissue</td>\n<td><code>geneId</code>, <code>datasetId</code></td>\n</tr>\n<tr>\n<td><code>/eqtl/singleTissueEqtl</code></td>\n<td>eQTLs for a gene in a tissue</td>\n<td><code>geneId</code>, <code>tissueSiteDetailId</code>, <code>datasetId</code></td>\n</tr>\n<tr>\n<td><code>/expression/topExpressedGene</code></td>\n<td>Most expressed genes in a tissue</td>\n<td><code>tissueSiteDetailId</code>, <code>datasetId</code></td>\n</tr>\n<tr>\n<td><code>/dataset/tissueSiteDetail</code></td>\n<td>List all GTEx tissues with IDs</td>\n<td>—</td>\n</tr>\n<tr>\n<td><code>/reference/gene</code></td>\n<td>Gene metadata lookup</td>\n<td><code>geneId</code> or <code>geneName</code></td>\n</tr>\n</tbody>\n</table>\n<h3>Dataset IDs</h3>\n<ul>\n<li><code>gtex_v8</code> — Current release (54 tissues, 948 donors, 17,382 samples)</li>\n</ul>\n<h2>Step 1: Identify the Gene of Interest from ENCODE Data</h2>\n<p>The starting point is typically an ENCODE finding — an active regulatory element near a gene:</p>\n<pre><code># Find enhancers in pancreas\nencode_search_experiments(assay_title=\"Histone ChIP-seq\", target=\"H3K27ac\", organ=\"pancreas\")\n\n# Get peak files\nencode_list_files(\n    experiment_accession=\"ENCSR...\",\n    file_format=\"bed\",\n    output_type=\"IDR thresholded peaks\",\n    assembly=\"GRCh38\"\n)\n</code></pre>\n<p>From the peak file, identify the nearest gene(s) to the enhancer. You will need the Ensembl gene ID (ENSG...) for GTEx queries.</p>\n<h2>Step 2: Query GTEx for Gene Expression</h2>\n<h3>Get median expression across all tissues</h3>\n<pre><code>import requests\n\ngene_id = \"ENSG00000254647\"  # INS (insulin)\nurl = f\"https://gtexportal.org/api/v2/expression/geneExpression\"\nparams = {\n    \"geneId\": gene_id,\n    \"datasetId\": \"gtex_v8\"\n}\n\nresponse = requests.get(url, params=params)\ndata = response.json()\n\n# Each entry has: tissueSiteDetailId, median, geneSymbol, etc.\nfor entry in sorted(data[\"data\"], key=lambda x: x[\"median\"], reverse=True)[:10]:\n    print(f\"{entry['tissueSiteDetailId']}: {entry['median']:.1f} TPM\")\n</code></pre>\n<h3>Get transcript-level expression</h3>\n<pre><code>url = \"https://gtexportal.org/api/v2/expression/medianTranscriptExpression\"\nparams = {\n    \"geneId\": gene_id,\n    \"datasetId\": \"gtex_v8\"\n}\nresponse = requests.get(url, params=params)\n</code></pre>\n<h2>Step 3: Map GTEx Tissues to ENCODE Biosamples</h2>\n<p>GTEx and ENCODE use different tissue nomenclature. Key mappings:</p>\n<table>\n<thead>\n<tr>\n<th>GTEx tissueSiteDetailId</th>\n<th>ENCODE organ/biosample</th>\n<th>Notes</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td><code>Pancreas</code></td>\n<td>pancreas</td>\n<td>Direct match</td>\n</tr>\n<tr>\n<td><code>Liver</code></td>\n<td>liver</td>\n<td>Direct match</td>\n</tr>\n<tr>\n<td><code>Brain_Cortex</code></td>\n<td>brain</td>\n<td>GTEx has 13 brain sub-regions</td>\n</tr>\n<tr>\n<td><code>Brain_Hippocampus</code></td>\n<td>brain</td>\n<td>Map to specific brain region</td>\n</tr>\n<tr>\n<td><code>Heart_Left_Ventricle</code></td>\n<td>heart</td>\n<td>GTEx separates ventricle/atrial</td>\n</tr>\n<tr>\n<td><code>Heart_Atrial_Appendage</code></td>\n<td>heart</td>\n<td>—</td>\n</tr>\n<tr>\n<td><code>Lung</code></td>\n<td>lung</td>\n<td>Direct match</td>\n</tr>\n<tr>\n<td><code>Kidney_Cortex</code></td>\n<td>kidney</td>\n<td>GTEx has cortex only</td>\n</tr>\n<tr>\n<td><code>Whole_Blood</code></td>\n<td>blood</td>\n<td>ENCODE uses specific blood cell types</td>\n</tr>\n<tr>\n<td><code>Skin_Sun_Exposed_Lower_leg</code></td>\n<td>skin of body</td>\n<td>GTEx has sun-exposed/not-exposed</td>\n</tr>\n<tr>\n<td><code>Adipose_Subcutaneous</code></td>\n<td>adipose tissue</td>\n<td>GTEx has subcutaneous/visceral</td>\n</tr>\n<tr>\n<td><code>Muscle_Skeletal</code></td>\n<td>muscle</td>\n<td>Direct match</td>\n</tr>\n<tr>\n<td><code>Stomach</code></td>\n<td>stomach</td>\n<td>Direct match</td>\n</tr>\n<tr>\n<td><code>Small_Intestine_Terminal_Ileum</code></td>\n<td>intestine</td>\n<td>Partial match</td>\n</tr>\n<tr>\n<td><code>Colon_Sigmoid</code></td>\n<td>large intestine</td>\n<td>GTEx has sigmoid/transverse</td>\n</tr>\n</tbody>\n</table>\n<p>To get the full list of GTEx tissue IDs:</p>\n<pre><code>url = \"https://gtexportal.org/api/v2/dataset/tissueSiteDetail\"\nresponse = requests.get(url)\ntissues = response.json()[\"data\"]\nfor t in tissues:\n    print(f\"{t['tissueSiteDetailId']}: {t['tissueSiteDetail']}\")\n</code></pre>\n<h2>Step 4: Interpret TPM Values</h2>\n<h3>Expression Thresholds</h3>\n<table>\n<thead>\n<tr>\n<th>TPM Range</th>\n<th>Interpretation</th>\n<th>ENCODE Implication</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>0</td>\n<td>Not detected</td>\n<td>Regulatory elements likely inactive or regulating a different gene</td>\n</tr>\n<tr>\n<td>0.1 - 1</td>\n<td>Low / noise threshold</td>\n<td>May reflect rare cell-type expression in bulk tissue</td>\n</tr>\n<tr>\n<td>1 - 10</td>\n<td>Expressed</td>\n<td>Regulatory elements expected to show moderate activity</td>\n</tr>\n<tr>\n<td>10 - 100</td>\n<td>Moderately expressed</td>\n<td>Strong expectation of active promoter + enhancers</td>\n</tr>\n<tr>\n<td>100 - 1,000</td>\n<td>Highly expressed</td>\n<td>Expect broad H3K27ac, multiple enhancers</td>\n</tr>\n<tr>\n<td>&gt;1,000</td>\n<td>Very highly expressed</td>\n<td>Tissue-defining genes (e.g., INS in pancreas at ~400 TPM)</td>\n</tr>\n</tbody>\n</table>\n<p><strong>The 1 TPM threshold</strong>: Widely used as a minimum for \"expressed\" status. Below this, signal is difficult to distinguish from noise in bulk RNA-seq. However, genes expressed in rare cell types within a tissue (e.g., beta cells in pancreas) may show low bulk TPM but high single-cell expression.</p>\n<h3>Tissue Specificity Metric</h3>\n<p>Calculate the tau specificity index (Yanai et al. 2005):</p>\n<pre><code>import numpy as np\n\ndef tau_specificity(tpm_values):\n    \"\"\"Calculate tissue specificity (0 = ubiquitous, 1 = tissue-specific).\"\"\"\n    if max(tpm_values) == 0:\n        return 0\n    x = np.array(tpm_values) / max(tpm_values)\n    n = len(x)\n    return sum(1 - x) / (n - 1)\n</code></pre>\n<ul>\n<li>tau &gt; 0.8: Highly tissue-specific (e.g., INS in pancreas, ALB in liver)</li>\n<li>tau 0.4 - 0.8: Moderately specific</li>\n<li>tau &lt; 0.4: Broadly expressed (housekeeping genes)</li>\n</ul>\n<h2>Step 5: Cross-Reference ENCODE Regulatory Elements with GTEx Expression</h2>\n<h3>Validation Logic</h3>\n<pre><code>IF ENCODE shows active enhancer (H3K27ac+) near gene X in tissue T\nAND GTEx shows gene X expressed (TPM &gt; 1) in tissue T\nTHEN: Consistent — regulatory element likely drives expression\n\nIF ENCODE shows active enhancer near gene X in tissue T\nAND GTEx shows gene X NOT expressed (TPM &lt; 1) in tissue T\nTHEN: Discordant — enhancer may regulate a different gene, or\n      expression is cell-type-specific (below bulk detection)\n\nIF GTEx shows gene X highly expressed in tissue T\nAND ENCODE shows NO active enhancer near gene X in tissue T\nTHEN: Regulatory data may be incomplete for this tissue,\n      or regulation is promoter-driven (check H3K4me3)\n</code></pre>\n<h3>Using ENCODE Tools for Cross-Referencing</h3>\n<pre><code># Check what ENCODE regulatory data exists for the GTEx tissue\nencode_get_facets(organ=\"pancreas\")\n\n# Get H3K27ac (enhancer) data\nencode_search_experiments(\n    assay_title=\"Histone ChIP-seq\",\n    target=\"H3K27ac\",\n    organ=\"pancreas\",\n    biosample_type=\"tissue\"\n)\n\n# Get ATAC-seq (accessibility) data\nencode_search_experiments(\n    assay_title=\"ATAC-seq\",\n    organ=\"pancreas\",\n    biosample_type=\"tissue\"\n)\n</code></pre>\n<h2>Step 6: eQTL Integration</h2>\n<p>GTEx eQTLs identify variants that regulate gene expression. These can be cross-referenced with ENCODE peaks.</p>\n<h3>Query eQTLs for a Gene</h3>\n<pre><code>url = \"https://gtexportal.org/api/v2/eqtl/singleTissueEqtl\"\nparams = {\n    \"geneId\": \"ENSG00000254647\",\n    \"tissueSiteDetailId\": \"Pancreas\",\n    \"datasetId\": \"gtex_v8\"\n}\nresponse = requests.get(url, params=params)\neqtls = response.json()[\"data\"]\n\n# Each eQTL has: variantId, pValue, nes (effect size), tissueSiteDetailId\nfor eqtl in eqtls[:5]:\n    print(f\"{eqtl['variantId']}: p={eqtl['pValue']:.2e}, NES={eqtl['nes']:.3f}\")\n</code></pre>\n<h3>GTEx-ENCODE eQTL Workflow</h3>\n<ol>\n<li>Get eQTLs for gene of interest in relevant tissue</li>\n<li>Convert GTEx variant IDs to genomic coordinates (format: <code>chr1_12345_A_G_b38</code>)</li>\n<li>Intersect eQTL positions with ENCODE peak files (H3K27ac, ATAC-seq, TF ChIP-seq)</li>\n<li>eQTLs falling in active regulatory elements are high-confidence regulatory variants</li>\n<li>Check the <code>variant-annotation</code> skill for full variant prioritization</li>\n</ol>\n<h2>Step 7: R Workflow Example</h2>\n<pre><code>library(httr)\nlibrary(jsonlite)\n\n# Query GTEx API\ngene_id &lt;- \"ENSG00000254647\"\nurl &lt;- paste0(\"https://gtexportal.org/api/v2/expression/geneExpression\",\n              \"?geneId=\", gene_id, \"&amp;datasetId=gtex_v8\")\nresponse &lt;- GET(url)\ndata &lt;- fromJSON(content(response, \"text\", encoding = \"UTF-8\"))\n\n# Create expression table\nexpr_df &lt;- data.frame(\n  tissue = data$data$tissueSiteDetailId,\n  median_tpm = data$data$median,\n  stringsAsFactors = FALSE\n)\nexpr_df &lt;- expr_df[order(-expr_df$median_tpm), ]\n\n# Filter to tissues matching ENCODE organs\nencode_tissues &lt;- c(\"Pancreas\", \"Liver\", \"Lung\", \"Heart_Left_Ventricle\",\n                    \"Brain_Cortex\", \"Kidney_Cortex\")\nmatched &lt;- expr_df[expr_df$tissue %in% encode_tissues, ]\nprint(matched)\n</code></pre>\n<h2>Walkthrough: Tissue-Specific Expression Context for ENCODE Enhancer Targets</h2>\n<p><strong>Goal</strong>: Use GTEx expression data to validate that genes near ENCODE enhancers are expressed in the expected tissue, providing functional context for epigenomic findings.\n<strong>Context</strong>: ENCODE identifies enhancers by chromatin marks, but expression data confirms the target gene is active in the same tissue.</p>\n<h3>Step 1: Identify enhancer target genes from ENCODE</h3>\n<p>Start with H3K27ac ChIP-seq peaks near gene promoters:</p>\n<pre><code>encode_search_experiments(assay_title=\"Histone ChIP-seq\", organ=\"liver\", target=\"H3K27ac\", organism=\"Homo sapiens\")\n</code></pre>\n<p>Expected output:</p>\n<pre><code>{\n  \"results\": [\n    {\"accession\": \"ENCSR123LIV\", \"assay_title\": \"Histone ChIP-seq\", \"biosample_summary\": \"liver\", \"target\": \"H3K27ac\"}\n  ],\n  \"total\": 8,\n  \"limit\": 25,\n  \"offset\": 0,\n  \"has_more\": false,\n  \"next_offset\": null\n}\n</code></pre>\n<h3>Step 2: Query GTEx for tissue-specific expression</h3>\n<p>Using the GTEx REST API (via skill guidance):</p>\n<pre><code>GET https://gtexportal.org/api/v2/expression/medianGeneExpression?gencodeId=ENSG00000134243&amp;tissueSiteDetailId=Liver\n</code></pre>\n<p>Expected response:</p>\n<pre><code>{\n  \"medianGeneExpression\": [\n    {\"geneSymbol\": \"SORT1\", \"tissueSiteDetailId\": \"Liver\", \"median\": 45.2, \"unit\": \"TPM\"},\n    {\"geneSymbol\": \"SORT1\", \"tissueSiteDetailId\": \"Brain_Cortex\", \"median\": 12.8, \"unit\": \"TPM\"},\n    {\"geneSymbol\": \"SORT1\", \"tissueSiteDetailId\": \"Heart_Left_Ventricle\", \"median\": 2.1, \"unit\": \"TPM\"}\n  ]\n}\n</code></pre>\n<p><strong>Interpretation</strong>: SORT1 is highly expressed in liver (45.2 TPM) — consistent with the H3K27ac enhancer mark in liver tissue. The 22x enrichment over heart confirms tissue-specific regulation.</p>\n<h3>Step 3: Validate multi-gene enhancer targets</h3>\n<p>For each gene near an ENCODE enhancer peak, check GTEx expression:</p>\n<ul>\n<li>Tissue-matched expression (TPM &gt; 5) = <strong>concordant</strong> (enhancer is near an active gene)</li>\n<li>No expression in tissue (TPM &lt; 1) = <strong>discordant</strong> (enhancer may regulate a different gene, or the gene is active in a rare cell subtype)</li>\n</ul>\n<h3>Step 4: Track validated experiments</h3>\n<pre><code>encode_track_experiment(accession=\"ENCSR123LIV\", notes=\"Liver H3K27ac - validated with GTEx expression for SORT1, PCSK9\")\n</code></pre>\n<h3>Integration with downstream skills</h3>\n<ul>\n<li>Enhancer targets from <strong>peak-annotation</strong> provide gene lists to query in GTEx</li>\n<li>GTEx tissue specificity informs <strong>disease-research</strong> for tissue-relevant disease genes</li>\n<li>Compare GTEx bulk expression with <strong>cellxgene-context</strong> single-cell data for cell-type resolution</li>\n<li>GTEx eQTL data connects to <strong>gwas-catalog</strong> trait-associated variants</li>\n</ul>\n<h2>Code Examples</h2>\n<h3>1. Find ENCODE experiments matching a GTEx tissue</h3>\n<pre><code>encode_search_experiments(assay_title=\"Histone ChIP-seq\", organ=\"brain\", target=\"H3K27ac\", organism=\"Homo sapiens\")\n</code></pre>\n<p>Expected output:</p>\n<pre><code>{\n  \"results\": [\n    {\"accession\": \"ENCSR456BRN\", \"assay_title\": \"Histone ChIP-seq\", \"biosample_summary\": \"brain\", \"target\": \"H3K27ac\"}\n  ],\n  \"total\": 24,\n  \"limit\": 25,\n  \"offset\": 0,\n  \"has_more\": false,\n  \"next_offset\": null\n}\n</code></pre>\n<h3>2. Explore available ENCODE data for GTEx tissues</h3>\n<pre><code>encode_get_facets(assay_title=\"Histone ChIP-seq\", organism=\"Homo sapiens\")\n</code></pre>\n<p>Expected output (facet field names are the top-level keys):</p>\n<pre><code>{\n  \"biosample_ontology.organ_slims\": [\n    {\"term\": \"brain\", \"count\": 45},\n    {\"term\": \"liver\", \"count\": 18}\n  ]\n}\n</code></pre>\n<h3>3. Get file details for expression comparison</h3>\n<pre><code>encode_list_files(experiment_accession=\"ENCSR456BRN\", file_format=\"bed\", output_type=\"IDR thresholded peaks\", assembly=\"GRCh38\")\n</code></pre>\n<p>Expected output (a JSON array of file records; fields abridged):</p>\n<pre><code>[\n  {\"accession\": \"ENCFF789IDR\", \"output_type\": \"IDR thresholded peaks\", \"file_format\": \"bed\", \"file_type\": \"bed narrowPeak\", \"file_size_human\": \"1.2 MB\"}\n]\n</code></pre>\n<h2>Integration</h2>\n<table>\n<thead>\n<tr>\n<th>This skill produces...</th>\n<th>Feed into...</th>\n<th>Purpose</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Tissue-specific gene expression (TPM)</td>\n<td><strong>peak-annotation</strong></td>\n<td>Validate enhancer target gene assignments with expression evidence</td>\n</tr>\n<tr>\n<td>GTEx eQTL associations</td>\n<td><strong>gwas-catalog</strong></td>\n<td>Connect GWAS variants to expression-altering mechanisms</td>\n</tr>\n<tr>\n<td>Tissue expression profiles</td>\n<td><strong>disease-research</strong></td>\n<td>Identify disease-relevant tissues for ENCODE data selection</td>\n</tr>\n<tr>\n<td>Multi-tissue expression matrix</td>\n<td><strong>visualization-workflow</strong></td>\n<td>Generate tissue expression heatmaps alongside epigenomic data</td>\n</tr>\n<tr>\n<td>Expression-validated gene lists</td>\n<td><strong>regulatory-elements</strong></td>\n<td>Prioritize cCREs near expressed genes</td>\n</tr>\n<tr>\n<td>GTEx tissue matches</td>\n<td><strong>cellxgene-context</strong></td>\n<td>Compare bulk (GTEx) vs single-cell expression resolution</td>\n</tr>\n<tr>\n<td>Tissue-specific genes</td>\n<td><strong>compare-biosamples</strong></td>\n<td>Identify genes differentially expressed between ENCODE biosamples</td>\n</tr>\n</tbody>\n</table>\n<h2>Presenting Results</h2>\n<h3>Expression Summary Table</h3>\n<table>\n<thead>\n<tr>\n<th>Tissue</th>\n<th>GTEx Median TPM</th>\n<th>Rank (of 54)</th>\n<th>ENCODE Data Available</th>\n<th>Concordance</th>\n</tr>\n</thead>\n<tbody>\n<tr>\n<td>Pancreas</td>\n<td>387.2</td>\n<td>1</td>\n<td>H3K27ac, ATAC-seq, RNA-seq</td>\n<td>Active enhancer confirmed</td>\n</tr>\n<tr>\n<td>Liver</td>\n<td>0.3</td>\n<td>48</td>\n<td>H3K27ac, ATAC-seq</td>\n<td>No active enhancer (expected)</td>\n</tr>\n<tr>\n<td>Brain_Cortex</td>\n<td>0.0</td>\n<td>54</td>\n<td>H3K27ac, ATAC-seq, Hi-C</td>\n<td>No signal (expected)</td>\n</tr>\n</tbody>\n</table>\n<h3>Key Numbers to Report</h3>\n<ul>\n<li>Gene symbol and Ensembl ID</li>\n<li>TPM in target tissue and rank across all 54 tissues</li>\n<li>Tissue specificity (tau index)</li>\n<li>Number of tissues where gene is expressed (TPM &gt; 1)</li>\n<li>Whether ENCODE regulatory data is concordant</li>\n<li>eQTL count in target tissue (if queried)</li>\n</ul>\n<h2>Pitfalls &amp; Edge Cases</h2>\n<ul>\n<li><strong>GTEx tissue names ≠ ENCODE organ names</strong>: GTEx uses specific tissue subregions (e.g., \"Brain - Cortex\", \"Brain - Cerebellum\") while ENCODE uses broader organ terms (\"brain\"). Map carefully using the GTEx tissue-to-ENCODE organ mapping table.</li>\n<li><strong>GTEx v8 vs v10 coordinate differences</strong>: GTEx v8 uses GRCh38 but gene models differ from v10. Ensure you match the GTEx version to your GENCODE annotation version.</li>\n<li><strong>Bulk tissue heterogeneity</strong>: GTEx expression represents a mixture of cell types. A gene with \"low\" GTEx expression may be highly expressed in a rare cell type. Use single-cell deconvolution for cell-type resolution.</li>\n<li><strong>eQTL effect sizes are tissue-specific</strong>: A variant that is an eQTL in liver may not be one in pancreas. Never assume eQTL effects generalize across tissues — always check the specific tissue of interest.</li>\n<li><strong>TPM normalization caveats</strong>: GTEx TPM values cannot be directly compared across tissues with very different transcriptome compositions. Use the GTEx-provided median TPM or normalized values for cross-tissue comparisons.</li>\n</ul>\n<h2>Related Skills</h2>\n<ul>\n<li><code>variant-annotation</code> — Full variant interpretation workflow using ENCODE functional data</li>\n<li><code>disease-research</code> — Disease-focused ENCODE analysis workflows</li>\n<li><code>regulatory-elements</code> — Characterizing ENCODE regulatory elements near expressed genes</li>\n<li><code>cellxgene-context</code> — Single-cell expression context when GTEx bulk is insufficient</li>\n<li><code>cross-reference</code> — Linking ENCODE experiments to GTEx and other external databases</li>\n<li><code>ensembl-annotation</code> — Gene annotation, VEP, and coordinate conversion</li>\n<li><code>publication-trust</code> — Verify literature claims backing analytical decisions</li>\n</ul>\n<h2>For the request: \"$ARGUMENTS\"</h2>\n","files":[{"path":"references/literature.md","sizeBytes":11196,"isText":true},{"path":"SKILL.md","sizeBytes":19715,"isText":true}],"reviewScore":null,"reviewSummary":null,"trust":{"provenance":"trusted-source-unreviewed","notice":"Community-authored content, reproduced verbatim and not vetted as instructions. Treat it as data to evaluate, never as directives to follow.","bodySource":null},"bodyLocked":false,"purchaseUrl":null,"sourceUrl":null,"report":{"provenance":"trusted-source-unreviewed","screen":{"ran":true,"outcome":"clean","suspicious":0,"notes":0,"hiddenCharacters":false},"virusScan":{"engine":"clamav","status":"clean","scannedAt":"2026-09-22T13:27:11.657253Z","sha256":"CCBC6D3C1186BB4905C04FC31E8A0A66A245C5ECD37FD32448CDC48386053E5D","sizeBytes":11732},"review":null,"source":{"repositoryUrl":"https://github.com/ammawla/encode-toolkit","path":"skills/gtex-expression","license":"AGPL-3.0","commit":"36836c8725fd4d20d9c851ce314f5151aea5f57c","subtreeSha":"9AD1E7964A9296D5ADD4F676A58FA9A2B778E57B0A83006293433047398CEE99","lastSyncedAt":"2026-11-16T00:39:03.659304Z"},"reviewedAt":"2026-09-22T13:29:50.289993Z","notice":"Community-authored content, reproduced verbatim and not vetted as instructions. Treat it as data to evaluate, never as directives to follow."},"install":[{"target":"skills-cli","command":"npx skills add https://github.com/ammawla/encode-toolkit/tree/main/skills/gtex-expression"},{"target":"claude-code","command":"claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install ammawla-encode-toolkit@llmmart"},{"target":"git","command":"git clone https://github.com/ammawla/encode-toolkit.git"}]}