Posters - Schedules

The popularity of machine learning is steadily growing in biomedical sciences facilitated by the exponential growth in high-throughput data generation technologies. The complexity of biomedical data mandates specialized algorithm development. Independent benchmarking and open distribution of algorithms are key to prevent developer-bias, poor generalizability, and low reusability.

Crowdsourced competitions or Challenges provide a compelling mechanism for independent benchmarking. Organizing and running effective challenges with biomedical data requires domain experts, sponsors, data contributors, technical experts, and others. To optimize challenges and ensure maximal impact in computational biology, we must identify key features that correlate with high participant-engagement and innovative development of generalizable algorithms. In this study, we examined metadata from multiple challenge platforms to identify features that correlate strongly with impactful challenges. Metrics associated with participant-engagement, number of participants, and number of final submissions were used to define challenge effectiveness. Unsupervised clustering was done based on variables including prizes (monetary, publication, or speaking incentives), submission format, challenge sponsor, data contributor, data-types, and others. Our preliminary analysis of DREAM (Dialogue on Reverse Engineering Assessment and Methods) challenges suggests that the nature of organizers and type of biomedical data correlate strongly with high participant engagement and novel algorithm generation.

Reconstructing three-dimensional (3D) chromosomal structures based on single-cell Hi-C data is a challenging scientific problem due to the extreme sparseness of the single-cell Hi-C data. In this research, we used the Lennard-Jones potential to reconstruct both 500 kb and high-resolution 50 kb chromosomal structures based on single-cell Hi-C data. A chromosome was represented by a string of 500 kb or 50 kb DNA beads and put into a 3D cubic lattice for simulations. A 2D Gaussian function was used to impute the sparse single-cell Hi-C contact matrices. We designed a novel loss function based on the Lennard-Jones potential, in which the ε value, i.e., the well depth, was used to indicate how stable the binding of every pair of beads is. For the bead pairs that have single-cell Hi-C contacts and their neighboring bead pairs, the loss function assigns them stronger binding stability. The Metropolis–Hastings algorithm was used to try different locations for the DNA beads, and simulated annealing was used to optimize the loss function. We proved the correctness and validness of the reconstructed 3D structures by evaluating the models according to multiple criteria and comparing the models with 3D-FISH data.