Research, academic papers, education, and exam prep (4.5K skills)
Finds claims in a biomedical paper that need citations, searches PubMed for relevant articles, and adds PMID markers to a new Markdown copy. It also creates a separate file of candidate references for later formatting with a reference manager.
Create a baseline characteristics section for a meta-analysis from supplied study data. The skill produces a descriptive paragraph and a formatted Markdown table in Chinese or English, then combines them into a consistent results section with a table reference.
Draft a structured Methods section for a meta-analysis using the supplied PICOS, eligibility criteria, risk-of-bias tool, and language. It covers the search strategy, screening, quality assessment, data extraction, and statistical analysis.
Turn medical notes, textbook excerpts, lecture summaries, or Q&A material into focused Anki cards. The workflow produces an importable TSV or card blueprint, states assumptions, and flags source material that is incomplete or too ambiguous to convert safely.
Shape a biomedical study idea into a realistic plan based on available samples, data, time, budget, lab capacity, and team skills. The skill identifies the main constraints, narrows the study route, defines a minimum executable version, and flags dependencies and assumptions.
Creates structured bulk transcriptome research plans for cancer studies focused on programmed or regulated cell death. It compares four workload levels and recommends a primary design, with analysis steps, figures, validation, a minimal plan, an upgrade path, and verified literature or a transparent search strategy.
Build a structured bioinformatics study plan for non-oncology diseases using bulk transcriptome data and a process-related gene theme. Compare four workload levels and define analysis steps, validation logic, figures, a minimal plan, and a publication upgrade path.
Creates a structured comparative network toxicology study plan for two related exposures and one shared toxic outcome. It maps shared and exposure-specific targets, proposes enrichment and network analyses, and lays out four workload levels, validation options, figures, and a cautious evidence-based interpretation.
Organize research team contributions to support author-order decisions, assign CRediT contributor roles, and prepare journal-ready statements. The skill can also flag possible inequities and provide recommendations for discussing authorship in collaborative projects.
Plan a bulk-transcriptome tumor biomarker study guided by immune infiltration. The skill structures analysis from differential expression and immune estimates through feature selection and diagnostic modeling, with four workload levels, validation options, and a figure plan.
Pressure-test a biomedical study design before data collection by classifying variable roles, identifying plausible confounders, and choosing how to control them. Get a reasoned adjustment strategy, likely bias risks, and mitigation steps, with uncertainty and residual bias made explicit.
Plan a computational study that compares gene expression across two diseases to identify shared biomarkers or hub genes. It lays out four workload options, recommends a primary workflow, and organizes analysis, validation, figures, and literature support.
Design a comparative bioinformatics study linking two related diseases through shared genes or biomarkers. It lays out four workload levels, recommends a primary plan, and organizes analyses such as enrichment, hub-gene selection, machine learning, immune interpretation, and validation with clear evidence limits.
Turn a biomedical study population into specific inclusion and exclusion rules that reviewers can apply consistently. Connect criteria to available records, tests, procedures, treatment history, and time windows, while flagging ambiguous cases and assumptions that need clarification.
Refine a broad biomedical research aim into a bounded, measurable objective that can guide study design. Identify the objective type, missing details, and exploratory versus confirmatory intent, then provide versions suited to plain-language, protocol, and downstream planning needs.
Plan locus-level studies that compare GWAS findings with eQTL, pQTL, sQTL, or related molecular signals. Get a structured workflow, candidate prioritization, workload options, and clear limits on what colocalization evidence can establish.
Plan validation before a medical study begins by matching the study’s main claim and available resources to appropriate validation layers. The skill distinguishes internal, external, temporal, and functional evidence, and clarifies what conclusions each can support.
Plan a two-sample Mendelian randomization study using a disease outcome and one or more exposures. The skill lays out instrument selection, harmonization, primary and sensitivity analyses, validation, and publication upgrades across four workload levels, with conservative interpretation and verified references.
Develop an assumption-aware sample size and power planning memo for clinical or translational studies. Identify the design and endpoint that drive planning, assess uncertain inputs, compare feasible scenarios, and outline fallback options without inventing precision when key assumptions are missing.
Plan an observational study of disease and biomarker associations using NHANES-style survey data. Get four workload options, a recommended plan, analysis steps, figures, validation boundaries, and a verified-reference search strategy tailored to the research question.
Build a structured study plan for prognostic biomarkers, from cohort and follow-up design to endpoint selection, marker screening, risk modeling, and validation. It helps researchers distinguish prognosis from diagnosis or treatment-response prediction and surface key feasibility and bias risks.
Turn one or more already identified biomedical research gaps into a traceable study plan. The output recommends one primary protocol, maps each gap to design choices, and lays out workflow, figures, validation, a minimal version, and an upgrade path without inventing references.
Plan a biomedical study that combines clinical variables with transcriptomics, proteomics, metabolomics, epigenomics, or related data. Define the use case, align data layers, choose feature reduction and modeling routes, and set validation steps across four workload levels.
Turn a clinical research question into a structured observational study plan using EHR, claims, or registry data. Define the cohort, index date, exposures, outcomes, follow-up, confounding approach, and assumptions the data cannot yet verify.