Welcome to the landing page for the hackathon. The competition is now over and final scores are now posted.
Please make sure each individual competing on your team is fully registered. To receive a prize, you must supply your University of Rochester e-mail address. All teams scoring better than random will receive a participation prize. 1st and 2nd place winning teams in each division will get a cash prize (see below). All team members must submit their own registration form to receive prizes. To finish your team registration or to register for an existing team, use this google form.
Hackathon-Summer-2024
,
owned by the team captain, will
be queried for a file named prediction/prediction.csv. Your predictions should be formatted in the exact same way and in the same order as the sample prediction file. Just replace the example TRUE/FALSE predictions with your own. If the team captain forks this
repository and writes predictions there everything should work
(as long as the predictions are formatted correctly). The scoreboard will be located here.
We cannot provide support beyond the diagnostic output included on the scoreboard if an error is encountered in scoring your predictions.
Single-cell multiome (RNA + ATAC) sequencing is a recently developed technology platform that simultaneously profiles gene expression and chromatin accessibility from the same cell. Single-cell multiome can transform your understanding of biology, for example, estimating the association between enhancer and gene expression using single-cell multiome data. However, such multiome data have an overabundance of zeros due to dropout events where the mRNA or chromatin accessibility is undetected in a cell. Thus, typical correlation methods (like Pearson correlation and Spearman's rank correlation) showed a low power in estimating the associations between regulatory elements and gene expression. For an overview of zero inflation in single-cell data, see Jiang et al. (2022, PMID: 35063006). Accurate detection of associations between regulatory elements and gene expression remains a significant challenge, and no single method outperforms all others.
The goal of this hackathon is to develop your model to identify associations between chromatin accessibility and gene expression using single-cell multiome data. Chromatin accessibility (ATAC in this hackathon) represents a functional canalization of the epigenome by defining a repertoire of putative regulatory regions (peaks in this hackathon) across the genome, further shaping gene expression programs (RNA in this hackathon). For an overview of the relationship between Chromatin accessibility and gene regulation, see Klemm et al. (2019, PMID: 30675018).
Find a complete description of the dataset here and the data here.
Congratulations to the winning teams:
1st place Undergraduate division- The Browns
2nd place Undergraduate division- Capybara
1st place Open division - CookieMonster and DMMH tied
2nd place Open division- BASE
If your team's final MCC was >0, you will receive a prize!
In order to claim your prize, you will need to fill out a post-competition survey and prize information form.
Thank you for participating in this competition! We hope that you enjoyed the event. Please give us feedback so that we can improve future events. All teams that submit predictions may receive a participation certificate. Please e-mail alarracu at bio.rochester.edu for your certificate.
Fill out the post-competition survey.