MOSAIC Community Challenge: Strains #2
Clinical Strain Detection Challenge Winners & Their Methods
Clinical Strain Detection challenge: Deep Dive with Jeremiah Faith
Introduction to the Clinical Strains Detection challenge
To develop biotherapeutic products based on live strains of commensal bacteria, researchers must have the ability to accurately detect and track how these product strains perform in patients after they are administered. Given that there may be closely related strains within a sample, the accurate detection of specific strains is the critical first step towards characterization of biotherapeutic products in terms of their dosage and efficacy.
To speed the translation of microbiome science into novel products, the Strains #2 Challenge aims to benchmark and advance computational methods for targeted strain detection, tracking the presence of certain known strains in a sample. Participants will contribute to the improvement of strain-level microbial analysis methods while also testing industry or proprietary tools in a secure and collaborative environment.
Insights from the Challenge will provide an objective comparison of the performance of different tools. Participants can submit multiple entries and see immediate results of their performance throughout the Challenge using the Mosaic Platform.
- Create Mosaic account and receive $50 worth of compute credits if you do not have a DNAnexus account yet!
- Read through the Rules of Participation and accept the Challenge.
Work on the Challenge by either:
Learn more about submitting your results.
- Creating your app on Mosaic, run it FOR FREE and submit results in the required format (and if you choose to share your method), or,
- Downloading the datasets, working outside Mosaic and submitting the results in the required format
The submissions to the Strains #2 Challenge will be evaluated using the Adjusted Rand Index. See Evaluation metrics section for a description of additional metrics.
- Challenge launch: February 12, 2018, 2:00 p.m. ET
- Challenge conclusion: August 3, 2018, 7:59 p.m. ET (1:59 a.m. CET)
- Results announced: Summer 2018
The microbiome industry is gearing up to introduce microbial-based products. Profiling a microbiome sample is the fundamental step in any microbiome analysis, but becomes even more challenging when higher resolution really matters. Current state-of-the-art methods based on next-generation sequencing are able to predict organisms down to the genera and species level. However, within the same species, individual strains can elicit different functionality and metabolic response, making strain-level identification and quantification key to the development and tracking of microbiome-based health solutions. It is critical to be able to accurately determine the type and quantity of microbes in a sample at the strain level in order to bring safe and effective products to market, and to accurately monitor their status within the human body. Participants will contribute to the improvement of sequencing standards while also testing their own tools and methods in a secure and collaborative environment.
- Recognition and benchmarking of methods
- Improved usability to drive citations
- Validation, scalability and robustness to drive industry collaboration
- Validated methods and pipelines that meet translational needs
- Access to “drop-in” apps that can be deployed into R&D pipelines immediately
- Seamless portability into secure, private, and compliant GXP environment for clinical research
- Community validation of methods to be used for submissions
- Reproducibility analyses
- Sustained usability to drive citation
Janssen Research & Development, LLC provides the vision, strategy, and design of Mosaic community challenges to seek answers to specific questions that are intended to speed the translation of science into novel products.
Mosaic has been developed and is maintained by DNAnexus Inc., a leader in cloud bioinformatics. DNAnexus has extensive experience in building secure and scalable platforms that enable computational analyses of biological data on the cloud.