By IU Simon Comprehensive Cancer Center
Nov 8, 2024
Cohort building simplified with BC²’s new data system
In this ongoing monthly feature, we’ll explore Biospecimen Collection and Banking Core's continuous efforts to integrate data pipelines, turning disparate data sources into a unified, accessible, and user-friendly data system for researchers.
Building on the previous update, faster feasibility assessments are on track to become a feature this year with BC²’s new data system in place. To make this possible, the team is now piloting an advanced tool: Manifold’s Cohort Explorer.
This easy-to-use data exploration tool empowers analysts and researchers to build cohorts, assess feasibility of studies, cross-tab multiple variables, and analyze data through an AI-powered natural language interface. Users can describe their analyses in plain English, and the Cohort Explorer’s AI handles the heavy lifting—rapidly filtering and analyzing complex datasets.
The Cohort Explorer will help simplify complex queries, making it easier to fulfill specific data and biospecimen requests that drive research projects forward. Here are some real examples of the types of requests that will be faster and easier to deliver with the Cohort Explorer:
- Treatment-naive lung adenocarcinoma cases with KEAP1, KRAS, and STK11 mutations
- Frozen tissue samples with TP53 mutations
- Cases with PIK3CA mutations
Key features of the Cohort Explorer
Natural language cohort building
Building cohorts is made easier with Cohort Explorer. By defining participant characteristics, treatment histories, or outcomes in plain English, researchers can create precise cohorts with ease. The tool’s intuitive design allows users to fine-tune parameters without advanced technical skills.
Effortless feasibility assessment
Cohort Explorer enables researchers to quickly assess the feasibility of their hypotheses. Using natural language, researchers can describe the outcomes, exposures, biospecimens and genomic data they wish to explore, with the AI dynamically generating results. This feature allows efficient validation of ideas without the need for manual data wrangling.
Insight into biobank inventory
Researchers and analysts will have a view of available biobank inventory including biospecimens, pre-indexed clinical data, genomic sequencing, variant calls, and annotations. This view will allow them to immediately understand what data is available to answer their initial research questions, in one single location.
Invitation to join the pilot program
Later this month, the first group of users will be invited to test the new system and provide feedback on how it will enhance current research workflows. Those interested in participating in the initial rollout can contact Jill Henry at jihenry@iu.edu or 317-278-2829 for access credentials.
