Can Geniml help with single-cell analysis?

Yes, it includes the scEmbed capability specifically designed for training embeddings on single-cell ATAC-seq data and integrating with scanpy workflows.

Does this skill support metadata-aware queries?

Yes, through the BEDspace module, Geniml can train shared embeddings for region sets and metadata labels to allow searching regions by label or vice versa.

What are 'universes' in the context of this skill?

Universes are reference peak sets built from collections of BED files using statistical methods like Maximum Likelihood or HMM to standardize genomic regions.

What is the primary use of the Geniml skill?

Geniml is used to apply machine learning models to genomic interval data, specifically for creating embeddings of BED files to enable similarity searches and clustering.

Geniml Genomic Machine Learning

Name: Geniml Genomic Machine Learning
Author: x-cmd

byx-cmd

•

データサイエンスとML

Applies unsupervised machine learning models to genomic interval data for embedding, clustering, and bioinformatics analysis.

Geniml is a specialized skill for analyzing and modeling genomic interval data (BED files) using machine learning. It enables Claude to implement advanced workflows including Region2Vec for region embeddings, scEmbed for single-cell ATAC-seq analysis, and BEDspace for joint region-metadata embeddings. This skill is essential for bioinformaticians building searchable genomic databases, standardizing consensus peak universes, or performing large-scale genomic feature learning within their development environment.

主な機能

01Automate BED file tokenization and embedding evaluation metrics

02Create joint embeddings for regions and metadata via BEDspace

03Analyze single-cell ATAC-seq data with scEmbed and scanpy integration

04Construct statistically rigorous consensus peak universes

058 GitHub stars

06Train unsupervised Region2Vec embeddings for genomic region sets

ユースケース

01Building cross-modal search engines for genomic regions and experimental labels

02Clustering and dimensionality reduction of single-cell chromatin accessibility data

03Standardizing genomic features across large-scale multi-experiment datasets

主な機能

01Automate BED file tokenization and embedding evaluation metrics

02Create joint embeddings for regions and metadata via BEDspace

03Analyze single-cell ATAC-seq data with scEmbed and scanpy integration

04Construct statistically rigorous consensus peak universes

058 GitHub stars

06Train unsupervised Region2Vec embeddings for genomic region sets

ユースケース

01Building cross-modal search engines for genomic regions and experimental labels

02Clustering and dimensionality reduction of single-cell chromatin accessibility data

03Standardizing genomic features across large-scale multi-experiment datasets