调研与学习技能

Run leave-one-out sensitivity analysis for a meta-analysis from a CSV file. Recalculate the pooled effect after removing each study in turn, then produce a forest plot and a data table to help assess how individual studies affect the overall result.

中文简介正在准备中。当前显示英语简介。

• Checking whether one study strongly affects a meta-analysis result
• Creating a leave-one-out forest plot from study data
• Exporting sensitivity results as a CSV table

Screen papers by comparing their titles and abstracts with your inclusion and exclusion criteria. Get a consistent Yes, No, or Maybe decision with a reason, ready to review when filtering studies for a systematic review or meta-analysis.

中文简介正在准备中。当前显示英语简介。

• Screening search results for a systematic review
• Filtering studies before a meta-analysis
• Applying inclusion and exclusion criteria consistently

Assess the quality and potential bias of a case-control medical study using the Newcastle-Ottawa Scale (NOS). Review selection, comparability, and exposure criteria, explain judgments with evidence from the paper, and organize the results into a structured summary table.

中文简介正在准备中。当前显示英语简介。

• Assessing bias in a case-control study
• Evaluating the quality of a medical paper
• Summarizing NOS ratings in a table

Create a forest plot for a survival-data meta-analysis from a CSV of study names, hazard ratios, and 95% confidence intervals. The workflow supports R and Python scripts and produces a PNG plot and a CSV data table, with pooled estimates and heterogeneity statistics when available.

中文简介正在准备中。当前显示英语简介。

• Visualizing survival meta-analysis results
• Generating a forest plot from study hazard ratios
• Exporting the plot and analysis table for a report

Create a meta-analysis forest plot for binary outcomes using a CSV of study names, event counts, and group sample sizes. The skill outputs a PNG plot and CSV table, with a combined effect estimate and heterogeneity information when the analysis succeeds.

中文简介正在准备中。当前显示英语简介。

• Plotting pooled results for binary outcomes across studies
• Creating a forest plot from event counts and sample sizes
• Exporting meta-analysis results as PNG and CSV

Assess whether a proposed topic may support a new meta-analysis by searching PubMed for clinical trials and existing meta-analyses. The workflow compares available studies with prior reviews to identify whether newer evidence may make an update viable.

中文简介正在准备中。当前显示英语简介。

• Checking a proposed topic before starting a meta-analysis
• Finding whether prior meta-analyses already cover a question
• Assessing whether newer clinical trials may support an update

Turn a CSV of study-level ROB2 assessments into two PNG figures: a Traffic Light Plot showing each study across assessment domains, and a Summary Bar Plot showing the risk distribution by domain. The input must contain the documented study and rating columns.

中文简介正在准备中。当前显示英语简介。

• Visualizing ROB2 assessments across studies
• Preparing figures for a systematic review
• Summarizing risk ratings by assessment domain

Assess bias and applicability in clinical diagnostic accuracy studies with QUADAS-2. Review patient selection, the index test, reference standard, and study flow, supporting each judgment with a paper quotation and marking missing information as unclear.

中文简介正在准备中。当前显示英语简介。

• Assessing a diagnostic accuracy paper with QUADAS-2
• Reviewing risk of bias across the four study domains
• Checking whether a study applies to a target setting

Screen a paper’s full text against inclusion and exclusion criteria for a meta-analysis, then return a structured Include or Exclude decision with a standard reason. If a PMID is supplied, PubMed title and abstract metadata can also be checked.

中文简介正在准备中。当前显示英语简介。

• Screening a full-text paper for a meta-analysis
• Applying PICO-based inclusion and exclusion criteria
• Checking PubMed metadata when a PMID is available

Create a funnel plot from meta-analysis data in a CSV file and assess possible publication bias with Egger’s regression and Begg’s rank correlation tests. The workflow checks required columns and study count, then produces a plot and test results.

中文简介正在准备中。当前显示英语简介。

• Checking possible publication bias in a meta-analysis
• Creating a funnel plot from study-level CSV data
• Reviewing Egger and Begg test results

Draft inclusion and exclusion criteria for a systematic review or meta-analysis from a title or keywords. The skill organizes eligibility around PICO—population, intervention, comparator, and outcomes—plus study design, and checks that exclusion rules do not contradict the inclusion criteria.

中文简介正在准备中。当前显示英语简介。

• Drafting eligibility criteria from a review title
• Turning review keywords into PICO-based criteria
• Checking exclusion rules against inclusion criteria

Extracts population, intervention or exposure, comparator, outcomes, and study design (PI(E)COS) from a clinical study or meta-analysis title, then returns the fields in a consistent JSON structure.

中文简介正在准备中。当前显示英语简介。

• Extracting PI(E)COS elements from a clinical study title
• Structuring the key elements of a meta-analysis title
• Preparing title-based information for a research workflow

Combine results from multiple studies by calculating pooled effect sizes, such as odds ratios or mean differences. Assess variation between studies and possible publication bias, then produce forest and funnel plots to help summarize and inspect the evidence.

中文简介正在准备中。当前显示英语简介。

• Pooling effect sizes across clinical studies
• Comparing study results with a forest plot
• Assessing heterogeneity with I-squared

Create a Baujat plot from meta-analysis data to show how individual studies contribute to the pooled result and statistical heterogeneity. The skill exports a PNG plot and a CSV of study contributions, helping locate studies for further review.

中文简介正在准备中。当前显示英语简介。

• Visualizing study contributions in a meta-analysis
• Finding studies that contribute strongly to heterogeneity
• Reviewing potential outlier studies

Create interactive medical imaging quizzes and cases using X-ray, CT, MRI, or ultrasound images. Set question types and difficulty for medical students, residents, or board exam study, with findings and differential diagnoses where requested.

中文简介正在准备中。当前显示英语简介。

• Creating a quiz from imaging cases
• Preparing radiology board exam questions
• Studying X-ray, CT, MRI, or ultrasound findings

Parse protein structure files in PDB or mmCIF format with Bio.PDB, inspect chains and residues, calculate geometric measurements, find nearby atoms or residues, and extract or superimpose structures for comparison.

中文简介正在准备中。当前显示英语简介。

• Reading PDB or mmCIF structure files
• Measuring distances, angles, or dihedrals
• Finding residues near a selected atom

Use Biopython tools to examine sequence motifs, restriction enzyme sites, codon usage, and related bioinformatics data. The skill helps produce structured analysis results and supports genomic feature visualization with GenomeDiagram.

中文简介正在准备中。当前显示英语简介。

• Summarizing motifs across aligned sequences
• Finding restriction enzyme cut sites in DNA
• Calculating codon frequencies and basic sequence statistics

Read and write phylogenetic trees with Bio.Phylo, convert common tree formats, prune taxa, reroot by an outgroup, calculate distances and basic statistics, and create terminal or Matplotlib visualizations for analysis and figures.

中文简介正在准备中。当前显示英语简介。

• Converting Newick, NEXUS, and phyloXML tree files
• Pruning taxa or rerooting a tree at an outgroup
• Calculating pairwise distances and branch length summaries

Generate USMLE Step 1 or Step 2 CK-style clinical cases for study and exam practice. Cases can include patient history, examination findings, tests, a single-best-answer question, answer choices, explanations, and learning objectives in text, JSON, or Markdown.

中文简介正在准备中。当前显示英语简介。

• Practicing Step 1 questions on basic science topics
• Studying diagnosis and management with Step 2 cases
• Generating cases focused on a specific condition

Turn informal invention notes into a structured patent disclosure, identify missing technical details, and prepare or analyze related-patent searches. It compares key features and offers a preliminary novelty assessment and drafting directions.

中文简介正在准备中。当前显示英语简介。

• Structuring an invention description for a patent disclosure
• Preparing keywords and IPC suggestions for a patent search
• Comparing related patents with an invention’s key features

Prepare for presentations by generating challenging questions about methods, statistics, interpretation, limitations, and future work. It provides response frameworks and key points to address, based on an abstract or research topic and a specified audience.

中文简介正在准备中。当前显示英语简介。

• Preparing answers to likely questions about a research abstract
• Rehearsing a presentation for an expert or general audience
• Identifying questions about methods, statistics, or study limitations

Draft or revise a research paper’s Methods section with the details readers need to understand and reproduce the study. Organize it around IMRAD conventions, reporting checklists, statistical choices, ethics, and journal requirements.

中文简介正在准备中。当前显示英语简介。

• Drafting a reproducible Methods section
• Checking a manuscript against CONSORT, STROBE, or PRISMA
• Addressing reviewer requests for methodological detail

Drafts clear, patient-friendly clinical trial recruitment ads from study details. It organizes eligibility, procedures, risks, rights, and contact information while avoiding guaranteed benefits, pressure tactics, and misleading claims.

中文简介正在准备中。当前显示英语简介。

• Drafting an ad for a clinical trial
• Explaining eligibility and study visits in plain language
• Preparing recruitment copy for print, digital, or social media

Extract ten baseline characteristics from a clinical trial article. Given a PMID, the workflow checks PubMed for basic metadata; when that is incomplete or unavailable, it uses the supplied full text to produce structured JSON and marks unreported fields as null or empty.

中文简介正在准备中。当前显示英语简介。

• Extracting baseline details from a clinical trial article
• Organizing study and participant information from full text
• Checking a PMID and completing missing details from article text