Single Cell Trajectory Contest

The Pinello Lab at Massachusetts General Hospital, LISH, and 25+ other single cell labs across the world collaborated to execute the Single Cell Trajectory contest. Single cell trajectory inference methods are essential tools in the analysis of cellular dynamic processes. In the field, there are over 60 methods used for prediction and measurement, and as such, the field lacks common methods that can be used across labs. The goal of the contest was to produce a new Docker-based, plug-and-play method for predicting single cell trajectory that surpassed identified benchmarks. While the focused outcome of the contest was the Docker solution, the process of organizing a community to converge on a problem statement, scoring criteria and contest benchmark also proved valuable, as it provided insight on how heterogenous communities can organize to set standards across a field of research. The contest resulted in 337 submissions from 36 competitors. In the first 48 hours, the solver submissions outperformed the benchmark and the winning submission resulting in a 14% increase over the benchmark. The contest was hosted on Topcoder with a prize pool of $24,000.