# Intro to ADNI and adnimerge for AD/PD Researchers **URL:** \ **Category:** Accessing and Understanding Data\ **Tags:** data-sharing, resources, neurodegenerative, data-access\ **Created:** [May 1, 2025, 1:24pm UTC](https://rcop.michaeljfox.org/t/intro-to-adni-and-adnimerge-for-ad-pd-researchers/993 "2025-05-01T13:24:42Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![hirotaka](https://sea1.discourse-cdn.com/michaeljfox/user_avatar/rcop.michaeljfox.org/hirotaka/32/58_2.png) [@hirotaka](https://rcop.michaeljfox.org/u/hirotaka)\ **Post date:** [May 1, 2025, 1:24pm UTC](https://rcop.michaeljfox.org/t/intro-to-adni-and-adnimerge-for-ad-pd-researchers/993/1 "2025-05-01T13:24:42Z")
## ADNI & `adnimerge` for AD/PD Researchers If you’re familiar with PPMI or other datasets hosted at [LONI](https://adni.loni.usc.edu/), you’ve likely come across the ADNI dataset as well. While ADNI is primarily designed for Alzheimer’s research, its rich longitudinal imaging, biomarker, and genetic data make it highly valuable for AD/PD comparative studies. I’ve used it myself for that purpose—for example, in [this CSF biomarker GWAS across PD and AD](https://pmc.ncbi.nlm.nih.gov/articles/PMC11459375/). For those new to ADNI’s structure, the R package `adnimerge` is a great starting point. It consolidates commonly used ADNI tables into a clean, analysis-ready format—eliminating the need for manual merges. The package is hosted at LONI, and [this page provides instructions](https://adni.bitbucket.io) on how to use it. A few important notes: - ADNI spans multiple phases (ADNI-1, GO, 2, 3). Some participants continue across phases, while others are newly recruited. This matters for longitudinal analyses. - The downloaded data files don’t always include full variable descriptions. I recommend using the [ADNI Data Dictionary Search](https://adni.loni.usc.edu/data-samples/data-dictionary-search/) for complete documentation. Happy to discuss if you’re working with ADNI or integrating it with PD datasets—feel free to share your experience or tips! ---
**Author:** ![ehutchins](https://sea1.discourse-cdn.com/michaeljfox/user_avatar/rcop.michaeljfox.org/ehutchins/32/530_2.png) [@ehutchins](https://rcop.michaeljfox.org/u/ehutchins)\ **Post date:** [June 3, 2025, 4:15pm UTC](https://rcop.michaeljfox.org/t/intro-to-adni-and-adnimerge-for-ad-pd-researchers/993/2 "2025-06-03T16:15:43Z")
Thank you for sharing this useful info, @hirotaka ! We often compare AD data to PD data for analyses so it is helpful to know more about this resource. Do you have any tips or tricks for data harmonization between ADNI and PD cohorts such as PPMI? @cameronreidhamilton @vcatterson the adnimerge package seems similar to the goals with the [Parkinson’s Insight Engine](https://github.com/MJFF-ResearchCommunity/PIE) (community [post here](https://rcop.michaeljfox.org/t/parkinsons-insight-engine-pie/962)). Might be worth taking a look. ---
**Author:** ![vcatterson](https://sea1.discourse-cdn.com/michaeljfox/user_avatar/rcop.michaeljfox.org/vcatterson/32/500_2.png) [@vcatterson](https://rcop.michaeljfox.org/u/vcatterson)\ **Post date:** [June 3, 2025, 6:45pm UTC](https://rcop.michaeljfox.org/t/intro-to-adni-and-adnimerge-for-ad-pd-researchers/993/3 "2025-06-03T18:45:52Z")
Thanks for highlighting this! The `adnimerge` package does look similar in scope to PIE. I’ll dig into it in some more detail 🙏 ---
**Author:** ![vdardov](https://sea1.discourse-cdn.com/michaeljfox/user_avatar/rcop.michaeljfox.org/vdardov/32/60_2.png) [@vdardov](https://rcop.michaeljfox.org/u/vdardov)\ **Post date:** [June 10, 2025, 3:18pm UTC](https://rcop.michaeljfox.org/t/intro-to-adni-and-adnimerge-for-ad-pd-researchers/993/4 "2025-06-10T15:18:16Z")
Thank you for sharing this! Very useful and appreciate you pointing to the R package and data dictionary.