Back to all discussions

Manager   •   about 2 years ago

IDEA/NEED: Helpful data sources for data science/AI/ML projects e.g. AACT database from CTTI

Hi everyone. For those interested in data science-driven AI/ML projects e.g. predicting clinical trial efficiency based on protocol characteristics, I wanted to open a thread so people understand potential resources they can try.

CTTI is the Clinical Trials Transformation Initiative, and they host a publicly-available database called AACT:
https://aact.ctti-clinicaltrials.org/
https://aact.ctti-clinicaltrials.org/snapshots

This includes content downloaded daily from ClinicalTrials.gov, and offers not only database access downloadable or in the cloud, but open source tools freely available, and source code via GitHub.

ClinicalTrials.gov is a potentially valuable resource for Track 2, in projects that may want to help guide designers of clinical trials to help predict and inform what will lead to higher success -- or may delay or decrease the chance of success -- of a clinical trial.

UPDATED: Here is a good article from a group called BCG that walks through how they analyzed the AACT database.
https://www.nature.com/articles/s41598-024-53211-z
The score and methods they derived are proprietary, but the AACT database can be freely downloaded from CTTI.

  • 1 comment

  •   •   almost 2 years ago

    That's a great one. I will like to work contribute to the development. Can you share with me the Github link, so i can see how i can contribute.

Comments are closed.