Skill UI

A persistent, structured knowledge base designed to accumulate and connect all facets of a research project—including published papers, initial ideas, conducted experiments, and testable scientific claims. Inspired by advanced LLM patterns, it builds a comprehensive knowledge graph, ensuring that research findings are cumulative and durable, rather than isolated notes. It manages typed relationships like contradiction, extension, and support to track the entire scientific lifecycle.

This technique guides the identification of musicians by prioritizing unique personal life events over discography details (albums or songs). It instructs users to use highly distinctive biographical information—such as career turning points, specific personal relationships, or unusual life incidents—as the primary search anchor. This method significantly improves search accuracy when typical musical clues are insufficient.

AnnData is a specialized Python package designed for managing and analyzing annotated matrices, particularly crucial in single-cell genomics and bioinformatics. It efficiently stores experimental measurements (X) alongside comprehensive metadata (obs, var, obsm, etc.). This skill allows users to read, write, subset, and concatenate large, complex datasets in formats like .h5ad and .zarr, making it essential for scRNA-seq analysis workflows and integration with the scverse ecosystem.

Provides programmatic access to massive, versioned public single-cell and spatial transcriptomics datasets from CZ CELLxGENE. This skill allows users to efficiently query specific subsets of cell metadata, gene expression matrices, embeddings, and summary statistics across various organisms, tissues, and diseases. It is ideal for large-scale, memory-efficient biological research and model training without full dataset downloads.

Generates professional, publication-ready clinical decision support (CDS) documents for pharmaceutical and clinical research settings. Capabilities include biomarker-stratified patient cohort analyses, evidence-based treatment recommendation reports (with GRADE grading), and advanced statistical analyses (e.g., hazard ratios, survival curves). Outputs are optimized for regulatory submissions and medical guidelines.

An autonomous research agent that plans complex investigations using Directed Acyclic Graphs (DAGs). It executes sub-questions in parallel, performs gap analysis, and iteratively refines findings before synthesizing a comprehensive, structured report. Ideal for deep market analysis, academic literature reviews, or complex problem-solving requiring multiple data sources.

DeepTools is a comprehensive command-line toolkit for processing and analyzing high-throughput sequencing data. It supports workflows for ChIP-seq, RNA-seq, ATAC-seq, and other NGS experiments. Core capabilities include converting BAM files to normalized coverage tracks (bigWig), performing quality control (QC), sample correlation analysis, and generating publication-quality visualizations, such as heatmaps and profile plots across genomic regions.

Geniml is a comprehensive Python package designed for applying machine learning models to genomic interval data, typically stored in BED files. It provides several unsupervised methods—including Region2Vec and BEDspace—to learn robust embeddings of genomic regions, single cells (scATAC-seq), and associated metadata. Use cases include dimensionality reduction, structural variation analysis, cell-type annotation, and cross-modal querying in genomics research.

Gtars is a high-performance toolkit written in Rust with Python bindings, designed for comprehensive analysis of genomic interval data. It provides specialized modules for overlap detection, generating coverage tracks (WIG/BigWig), tokenization for Machine Learning models, and detailed fragment analysis in single-cell genomics. Use it for robust bioinformatics pipelines and computational genomics research.

Molfeat is a comprehensive Python library designed for molecular featurization. It converts chemical structures (SMILES/RDKit) into numerical feature vectors, supporting over 100 featurizers like ECFP, MACCS, and deep embeddings (ChemBERTa). It unifies featurization pipelines, making it ideal for QSAR modeling, virtual screening, and integrating into deep learning ML workflows.

An end-to-end toolkit designed for analyzing high-density Neuropixels electrophysiology recordings. It implements best practices using SpikeInterface, covering the entire workflow: data loading (SpikeGLX, NWB), preprocessing, motion/drift correction, advanced spike sorting (Kilosort4), quality metric computation, and unit curation (including AI-assisted review). Ideal for neuroscience research.

UMAP (Uniform Manifold Approximation and Projection) is a powerful technique for non-linear dimensionality reduction. This skill enables the creation of scalable, structure-preserving embeddings (2D/3D) for visualization, clustering preprocessing, and advanced machine learning feature engineering. It supports various metrics (e.g., Euclidean, Cosine) and supervised learning workflows.