Skill UI

An expert agent designed to collect critical data assets—including metadata, data lineage, and query logs—from any data warehouse platform. It uses the push ingestion API to send this data directly to Monte Carlo. This method is essential for integrating data sources that cannot be reliably covered by standard pull collectors, such as custom lineage paths or historical query logs.

AnnData is a robust Python package designed for handling annotated data matrices, primarily used in single-cell genomics and multi-omics analysis. It efficiently stores experimental measurements (X) alongside comprehensive metadata (observation, variable, and various annotations). This skill allows users to create, manipulate, subset, and concatenate large, complex datasets, supporting formats like .h5ad and .zarr, and integrating seamlessly with the scverse ecosystem (Scanpy, scvi-tools).

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.

This skill provides programmatic access to the CZ CELLxGENE Census, a comprehensive, versioned repository containing over 61 million single-cell genomics records from human and mouse. It is essential for conducting large-scale cross-dataset analyses, querying gene expression data across diverse tissues, cell types, and diseases, and training machine learning models in bioinformatic research. It supports both in-memory and out-of-core computation using advanced libraries like scanpy.

Dask is a powerful Python library designed for scalable, parallel, and distributed computing. It enables workflows for datasets that exceed available RAM or require processing across multiple machines. Users can scale familiar pandas and NumPy operations (DataFrames, Arrays) and process unstructured data (Bags) efficiently, making it ideal for large-scale ETL, analytics, and distributed machine learning tasks.

DeepChem is a comprehensive Python library designed for applying machine learning to chemical, biological, and materials science datasets. It enables molecular property prediction (e.g., ADMET, toxicity), drug discovery, and materials design. The skill supports diverse featurization methods (fingerprints, descriptors, graph representations), handling various data formats like SMILES and SDF, and implementing advanced GNN models for complex scientific research.

Perform deep exploratory data analysis (EDA) across a vast array of scientific formats, including chemistry, bioinformatics, microscopy, and spectroscopy. This skill automatically detects the file type, assesses data quality and integrity, extracts format-specific metadata, and generates detailed markdown reports. It is ideal for understanding the structure, content, and potential analysis paths for complex scientific datasets.

GeoPandas extends pandas to enable powerful spatial operations on vector data. It is essential for tasks in Geographic Information Systems (GIS), allowing users to read, write, and analyze multiple formats like GeoJSON, Shapefile, and GeoPackage. Key functionalities include spatial joins, calculating geometric properties (e.g., area, buffer), reprojecting coordinates (CRS management), and creating interactive choropleth maps. Use it for advanced spatial analysis, such as overlay operations, dissolving boundaries, and nearest neighbor lookups across different datasets.

Histolab is a Python library designed for advanced whole slide image (WSI) processing in digital pathology. It automates complex tasks such as loading diverse WSI formats, sophisticated tissue segmentation (creating binary masks), and extracting high-quality, informative tiles. The library supports multiple tiling strategies (random, grid, score-based), making it ideal for preparing standardized datasets for deep learning models and accelerating pathological research pipelines.

Hypogenic is a powerful framework designed to accelerate scientific discovery by automating the process of hypothesis generation and testing on complex tabular datasets. It supports three advanced modes: HypoGeniC for purely data-driven pattern discovery; HypoRefine for synergistically integrating literature insights with empirical data; and Union methods for comprehensive hypothesis coverage. Ideal for domains like deception detection, content analysis, and advanced predictive modeling where systematic exploration of multiple theories is required.

LaminDB is an open-source, unified data framework designed specifically for complex biological datasets. It provides a robust lakehouse architecture that ensures data is queryable, traceable, reproducible, and FAIR (Findable, Accessible, Interoperable, Reusable). Use this skill for managing multi-modal genomic data (scRNA-seq, spatial transcriptomics), tracking computational lineage, curating data using standard biological ontologies, and integrating advanced ML/AI workflows.

A comprehensive toolkit for running and analyzing molecular dynamics (MD) simulations. This skill covers the entire workflow: system setup, force field definition, energy minimization, and equilibration using industry-standard tools like OpenMM and MDAnalysis. Essential for structural biology, drug binding studies, and biophysics research.