geniml

This skill should be used when working with genomic interval data (BED files) for machine learning tasks. Use for training region embeddings (Region2Vec, BEDspace), single-cell ATAC-seq analysis (scEmbed), building consensus peaks (universes), or any ML-based analysis of genomic regions. Applies to BED file collections, scATAC-seq data, chromatin accessibility datasets, and region-based genomic feature learning.

1.74x

Quality

71%

Does it follow best practices?

Impact

68%

1.74x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./scientific-skills/geniml/SKILL.md

Evaluation results

40%

Building Genomic Region Embeddings for ChIP-seq Data

Region2Vec pipeline with universe building and evaluation

Criteria

Without context

With context

uv install command

hard_tokenization used

p_value_threshold 1e-9

region2vec function used

25%

num_shufflings 1000

init_lr in range

embedding_dim specified

100%

Coverage validation

100%

Coverage threshold check

100%

evaluate_embeddings called

Multiple eval metrics

22%

55%

Pipeline order correct

100%

74%

48%

Single-Cell ATAC-seq Cell Clustering Pipeline

scEmbed single-cell ATAC-seq analysis with scanpy integration

Criteria

Without context

With context

AnnData var chr/start/end

100%

tokenize_cells used

Pre-tokenization before training

100%

ScEmbed class used

100%

scembed_X obsm key

100%

neighbors uses scembed_X

100%

leiden clustering

100%

UMAP computed

100%

uv install used

Filtered data note

scanpy integration

100%

92%

36%

Building a Searchable Genomic Region Database

BEDspace metadata-aware region search

Criteria

Without context

With context

Four-step workflow

100%

metadata file_name column

100%

StarSpace path parameter

100%

StarSpace install note

100%

Consistent universe file

37%

100%

r2l search type

100%

l2r search type

100%

r2r search type

100%

geniml bedspace commands

25%

100%

uv install used

labels parameter

25%

100%

README explains format

62%

100%

Repository: K-Dense-AI/claude-scientific-skills
Commit: b58ad7e

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Building Genomic Region Embeddings for ChIP-seq Data Single-Cell ATAC-seq Cell Clustering Pipeline Building a Searchable Genomic Region Database

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.