Skill UI

A practical decision framework for determining whether to use regular expressions or Large Language Models (LLMs) when parsing structured text (e.g., forms, quizzes, invoices). It recommends starting with deterministic, low-cost regex for 95-98% of structured data, and only invoking expensive LLM calls for identifying and validating ambiguous edge cases, optimizing both cost and accuracy.

Provides programmatic access to the CZ CELLxGENE Census (61M+ cells) for querying standardized single-cell expression, metadata, embeddings, and stats across tissues and diseases, supporting scanpy and ML workflows.

Programmatic access to the COSMIC cancer mutation database for downloading mutations, Cancer Gene Census, mutational signatures, fusions, copy number, expression, and resistance data to fuel oncology research or bioinformatics pipelines after authentication.

Connect to NCBI's GEO repository, search series/samples/platforms via Entrez queries, fetch profiles, and download raw/processed matrices using GEOparse for transcriptomics and expression analysis workflows.

PyDESeq2 brings the DESeq2 differential expression workflow to Python, letting users load bulk RNA-seq counts, specify design formulas, fit DESeq2 models, run Wald tests with FDR correction, optionally apply apeGLM shrinkage, and export gene-level statistics within pandas/AnnData pipelines.

Query the Reactome Content and Analysis Services (or the reactome2py client) to fetch pathway hierarchies, map genes, run overrepresentation/expression analyses, and explore disease interactions for systems biology research.

Probabilistic deep-learning framework built on PyTorch for single-cell omics; handles batch correction, integration, multimodal and spatial data, differential expression, and transfer learning with unified APIs.

Access multi-dimensional cancer genomics data via cBioPortal REST API for mutation, copy-number, expression, and survival insights across TCGA/GENIE studies, enabling target validation and clinical association analyses programmatically with Python clients.

Query the GTEx portal and API to retrieve tissue-specific gene expression, eQTL, sQTL, and variant-level regulatory data for interpreting GWAS loci and exploring gene regulation across 54 human tissues.

Arboreto is a powerful Python library designed for inferring Gene Regulatory Networks (GRNs) from various gene expression datasets, including bulk and single-cell RNA-seq data. It leverages advanced machine learning algorithms like GRNBoost2 and GENIE3. The library supports scalable, distributed computation using Dask, making it ideal for analyzing large-scale, complex transcriptomics datasets to identify critical transcription factor-target gene relationships.

Provides programmatic and standardized access to massive, versioned datasets of single-cell and spatial transcriptomics from CZ CELLxGENE. This skill allows users to query millions of cells, analyze gene expression, and access pre-calculated embeddings across various tissues, diseases, and species without downloading entire datasets. Essential for large-scale bioinformatics analysis, machine learning model training, and comprehensive cross-dataset comparison.

Polars is a lightning-fast DataFrame library built for Python and Rust, leveraging Apache Arrow. It provides an expression-based API and lazy evaluation framework for highly efficient data manipulation. It is ideal for complex data pipelines, large-scale ETL tasks, and migrating from Pandas due to its superior performance, parallel execution capabilities, and optimized query planning.