<!-- BSV — Brief Skill View | поиск: BSV Скил : archcode-genomics TL;DR : Патогенность геномных вариантов (биоинформатика) Вызов : 'геномный анализ', 'archcode' НЕ для : Клинические решения -->
Scanned 9/2/2026
Install to Claude Code
npx -y skills add sergeeey/Claude-cod-top-2026 --skill archcode-genomics --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Archcode Genomics?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/sergeeey-archcode-genomics)More formats (shields.io, HTML) on the badges page.
<!-- BSV — Brief Skill View | поиск: BSV
Скил : archcode-genomics
TL;DR : Патогенность геномных вариантов (биоинформатика)
Вызов : 'геномный анализ', 'archcode'
НЕ для : Клинические решения
-->
---
name: archcode-genomics
description: >
[STATUS: review] [CONFIDENCE: high] [REVIEWED: 2026-03-13]
USE when working with genomic data, variant analysis, or chromatin structure.
Triggers: ClinVar, chromatin, variant, extrusion, loop, locus, HBB, SpliceAI,
pathogenicity, genomics, ARCHCODE, structural variant, pearl.
paths: "**/*variant*,**/*vcf*,**/*clinvar*,**/*extrusion*,**/*chromatin*"
effort: max
---
# ARCHCODE Genomics Skill
## Project Context
ARCHCODE is a physics-based method for predicting pathogenicity of non-coding variants through simulation of chromatin loop extrusion. It detects "pearl variants" — variants invisible to sequence-based methods (VEP, CADD, SpliceAI).
## Key Data
- **9 validated loci:** HBB, HBA, BCL11A, CTCF-rich, GATA1, beta-globin LCR, SHH-ZRS, SOX9-KCNJ, MYC-TAD
- **# YOUR_VARIANT_COUNT ClinVar variants** analyzed
- **# YOUR_PEARL_COUNT pearl variants** detected (pathogenic, invisible to VEP/SpliceAI)
## Pearl Variant — Definition
A variant that:
1. VEP = benign/modifier (not pathogenic)
2. SpliceAI delta = 0 (no splicing effect)
3. MPRA enrichment = null (no functional data)
4. BUT: disrupts 3D chromatin structure (CTCF binding, loop anchor, TAD boundary)
## Pipeline
```
ClinVar VCF → VEP annotation → SpliceAI filtering →
ARCHCODE simulation (loop extrusion) →
Structural impact scoring → Pearl detection
```
## Competitors (for Discussion section)
- **ncVarPred-1D3D** (PMID 37669132) — sequence + 3D, but without ClinVar pearl detection
- **PRISMR** (PMID 29662163) — polymer modeling, but without clinical validation
- **Sei** — sequence-only, does not see structural blind spots
## MCP Servers
Switch to SCIENCE profile: `switch-profile.ps1 science`
- ncbi-datasets: gene info, sequences
- uniprot: protein features, variants, domains
- pubmed-mcp: literature search
## Gotchas
- VEP/CADD miss non-coding variants — always check 3D chromatin context
- Pearl variants require loop extrusion simulation — sequence-only methods fail
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!