Can I perform multi-treatment comparisons?

Absolutely. The skill includes a Network Meta-Analysis (NMA) consistency model specifically designed for comparing multiple treatments and generating treatment rankings.

How does the skill handle model convergence issues?

For random effects models, the skill utilizes non-centered parameterization in Stan, which is a recommended best practice to avoid sampling difficulties in 'funnel' geometries.

Does this skill support both Stan and JAGS?

Yes, it provides implementation templates for both Stan (using C++-like syntax) and JAGS (using BUGS-like syntax) to accommodate different statistical modeling preferences.

Can I model binary outcomes with this skill?

Yes, it includes specific templates for binary outcomes using log-odds ratios and binomial-logit likelihoods to compare events between treatment and control groups.

Bayesian Meta-Analysis Models

Name: Bayesian Meta-Analysis Models
Author: choxos

bychoxos

0•

Ciencia de Datos y ML

Provides comprehensive Bayesian meta-analysis templates using Stan and JAGS for advanced biostatistical evidence synthesis.

This skill enables users to implement sophisticated Bayesian meta-analysis techniques directly within their workflow. It covers a wide spectrum of models, from simple fixed and random effects to complex network meta-analyses (NMA) and publication bias selection models. By offering standardized code patterns in both Stan and JAGS, it facilitates robust statistical modeling, handling of binary outcomes via log-odds, and the calculation of critical metrics like heterogeneity (I²) and treatment rankings. It is an essential resource for biostatisticians and data scientists performing systematic reviews or evidence-based research.

Características Principales

01Network Meta-Analysis (NMA) consistency models with treatment ranking logic

020 GitHub stars

03Selection models to identify and adjust for publication bias

04Non-centered parameterization templates for improved MCMC convergence

05Automated calculation of heterogeneity, prediction intervals, and effect probabilities

06Stan and JAGS implementations for fixed and random effects models

Casos de Uso

01Adjusting research findings for potential publication bias in systematic reviews

02Comparing multiple medical interventions through network meta-analysis

03Synthesizing clinical trial data to determine overall treatment efficacy

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add choxos/biostatagent meta-analysis

For use in Claude.ai and ChatGPT

Download Skill