Skill UI

A comprehensive toolkit for securely inspecting and automating complex microscopy data workflows against OMERO.server. It provides deep programmatic access using omero-py, covering metadata, annotations, ROIs, rendering, and detailed data export/import planning. Essential for building controlled, large-scale data inventory and analysis pipelines.

PathML is a dedicated computational pathology software designed for local, research-only workflows. It enables users to load and tile whole slide images, build comprehensive preprocessing and quality control pipelines, quantify multiplex data, construct spatial graphs, and plan bounded model inference. It is crucial for researchers handling highly sensitive medical images while maintaining strict data privacy and local processing standards.

PyHealth is a comprehensive Python toolkit designed for building deep-learning pipelines in the clinical and healthcare domains. It supports diverse data modalities, including EHR records, physiological signals, and medical images. The framework follows a stable 5-stage pipeline (Dataset → Task → Model → Trainer → Metrics), enabling users to perform complex tasks like predicting mortality, optimizing drug recommendations, staging sleep, or cross-mapping medical codes (ICD/ATC/RxNorm) on datasets like MIMIC-III/IV and OMOP.

PyMC is a comprehensive Python library for probabilistic programming and Bayesian modeling. It enables users to build, fit, and validate complex statistical models, including hierarchical structures, time series, and regression models. It provides powerful tools for MCMC sampling (NUTS) and variational inference, making it essential for rigorous uncertainty quantification and model comparison (LOO/WAIC).

This skill compiles comprehensive external evidence required for drafting high-quality scientific manuscripts or research briefs. It goes beyond simple link lists, targeting verified, unique academic references (e.g., 60 references) and generating a structured research packet. Use it when you need literature reviews, evidence supporting a claim, or a deep dive into scientific background, methodology, or limitations.

A highly structured framework designed for generating, organizing, and transparently prioritizing scientific research concepts. It enforces independent idea generation, explicit assumption labeling, structured discussion, and evidence-aware evaluation, ensuring the creation of robust, proposal-level research directions for early-stage exploration.

This skill provides a systematic framework for evaluating the rigor of scientific claims, research methodology, and evidence quality. It guides users through assessing experimental design validity, identifying various types of biases (e.g., selection, cognitive), and applying established grading frameworks like GRADE or Cochrane Risk of Bias. Ideal for systematic reviews, peer review, and critical analysis of scientific literature.

A comprehensive guide and tooling mechanism for creating and auditing scientifically truthful, accessible, and publication-ready figures. It enforces non-negotiable guardrails, guiding users on data integrity, proper encoding (e.g., handling uncertainty, missing values), and accessibility standards (WCAG). Supports advanced plotting libraries like Matplotlib and Seaborn, focusing on methodological rigor over mere aesthetics.

scikit-bio is a comprehensive Python library designed for advanced bioinformatics and biological data analysis. It provides robust capabilities for sequence manipulation (DNA, RNA, Protein), performing complex sequence alignments (pairwise and multiple), constructing and analyzing phylogenetic trees, and calculating diverse ecological metrics. It is essential for microbiome analysis, evolutionary studies, and general biological data processing, supporting standard formats like FASTA, Newick, and BIOM.

A comprehensive resource detailing the use of scikit-learn, the industry-standard Python library for classical machine learning. Use this skill for the entire ML lifecycle, including supervised tasks (classification, regression), unsupervised tasks (clustering, dimensionality reduction), data preprocessing, model evaluation, and building production-ready pipelines.

A comprehensive skill for performing statistical survival analysis on time-to-event data, specifically handling right-censored outcomes and competing risks. It provides robust workflows for model building (including Cox PH, Coxnet, and deep learning integration), leakage-safe preprocessing, rigorous model selection, and evaluation metrics (e.g., discrimination, calibration). Ideal for medical, reliability, and failure analysis.

scVelo is a powerful Python package designed for RNA velocity analysis using single-cell RNA-seq data. It estimates cell state transitions by modeling the kinetics of mRNA splicing—specifically using the ratio of unspliced to spliced mRNA abundances. This allows researchers to reconstruct developmental trajectories and predict cell fates, even when only snapshot data is available, making it a vital complement to standard trajectory inference tools like Scanpy.