snaptron

Query Snaptron for splice junction counts, RNA-seq metadata, and BigWig coverage.

1|Updated Oct 25, 2020
One-click install
npx skills add https://github.com/bfairkun/dotfiles --skill snaptron-bfairkun
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: snaptron
Source: https://github.com/bfairkun/dotfiles/tree/main/agents/.agents/skills/snaptron
Command: npx skills add https://github.com/bfairkun/dotfiles --skill snaptron-bfairkun

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyBigWig, httpx.

What problem does it solve? Querying large-scale RNA-seq resources like SRA, GTEx, and TCGA for splice junction counts and per-base coverage requires navigating the Snaptron REST API's compilation-specific quirks, metadata field differences, and BigWig URL construction rules, which this Skill documents and operationalizes. ## Core Features & Use Cases - Sample Search: Find RNA-seq samples across compilations (srav2, srav3h, gtexv2, tcgav2, ccle, mouse) using metadata filters like cell_line, tissue, or study_title. - Junction & Coverage Queries: Retrieve splice junction counts per sample and per-base coverage from remote BigWig files via HTTP range requests. - Expression & Visualization: Compute normalized junction expression ratios, approximate TPM from BigWig exon coverage, and generate IGV session XML files. - Use Case: Compare MYT1L splicing in HeLa samples by searching srav2 for rail_ids, querying junction counts, building BigWig URLs, and saving an IGV session for visual inspection. ## Quick Start Ask the assistant to search Snaptron srav2 for HeLa samples and query GAPDH junction coverage for the top hits.

Frequently Asked Questions about snaptron

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I query splice junction counts from Snaptron?▼

Use the snaptron endpoint with a gene or region, e.g. GET https://snaptron.cs.jhu.edu/srav2/snaptron?regions=GAPDH, optionally filtered by sample rail_ids via sids. Results return as TSV with per-sample counts in the samples column.

How to search RNA-seq samples by cell line in Snaptron?▼

Use the samples endpoint with sfilter, e.g. sfilter=cell_line:HeLa on srav2. Note srav3h has no cell_line field; search study_title, study_abstract, or sample_description instead.

What is the difference between srav2 and srav3h compilations?▼

srav2 covers ~49K recount2 human samples with cell_line and run_acc fields, while srav3h covers ~316K recount3 samples but uses sample_acc and lacks cell_line. srav3h data freezes around March 2021 and has no bigwig_file field.

How do I build BigWig URLs for Snaptron samples?▼

For srav2, combine study_accession and run_accession: https://duffel.rail.bio/recount/{study}/bw/{run}.bw. For srav3h, use the recount3 base_sums path pattern with study and run suffix subdirectories.

Why does my srav3h ids query return broken results?▼

The MCP client's fetch_samples creates a broken pandas MultiIndex for srav3h ids= queries. Parse the TSV response directly with csv.reader instead, as shown in the manual metadata fetching example.

Can I visualize Snaptron coverage in IGV without downloading files?▼

Yes, IGV streams BigWigs directly from duffel.rail.bio URLs using HTTP range requests. Generate a session XML with Resource and Track elements pointing to the remote BigWig URLs and open it via File > Open Session.