Challenge Overview

Participants will be provided with information about a biological target, including a set of compounds with known activity and a blind dataset consisting of confirmed active compounds and decoys. The blind dataset will contain approximately 3,000 structures.

Objective

The goal of the challenge is to select 100 compounds from the blind dataset that maximize the number of true active hits within a fixed time limit of approximately 5.5 hours.

Methods

Participants are free to use any computational approach or software tools, including but not limited to:

• Similarity search
• Pharmacophore modeling
• QSAR modeling
• Molecular docking

Teams and Submissions

• Participants may form teams of up to three members
• Teams may include on-site and remote participants
• Each team may submit up to 10 predictions
• Only the best-performing submission per team will be considered for scoring

Evaluation and Awards

• On-site and remote teams will be evaluated separately
• A final leaderboard will identify the top three teams in each category which will be awarded by prizes provided by RowanSci
• The top three teams from both categories will be invited to prepare a short presentation (1–2 slides) describing their winning solution

Post-Challenge Prospective Validation

After the challenge concludes, predictions from the top-performing teams will be considered for compound purchasing and experimental validation.

Registration will be opened a week in advance.