Yiqun Chen and Kun Yue, University of Washington PhD students in the Department of Biostatistics, won students awards at the recent annual conference of the Western North American Region (WNAR) of the International Biometric Society.
Chen’s presentation, “Quantifying uncertainty in spikes estimated from calcium imaging data,” earned Best Student Oral Presentation honors.
Calcium imaging is an increasingly popular technique in the field of neuroscience that can image large populations of neurons simultaneously in living organisms and animals. The presentation proposed a framework to quantify the uncertainty associated with spikes estimated from calcium imaging data.
“Our work is positioned to have a direct impact on the analysis of real-world calcium imaging data, as neuroscientists use estimated spikes to understand how the brain processes information,” said Chen.
Yue was one of three students who tied for the Best Student Paper award with her work, “REHE: Fast Variance Components Estimation for Linear Mixed Models.”
Linear mixed models are widely used in ecological and biological applications, especially in genetic studies. Commonly used variance component estimators for linear mixed models are computationally inefficient in large samples, may be unstable with small samples, and/or may yield negative estimates of variances. Yue’s paper proposes regularized estimators along with an inference framework as fast and robust alternatives.
“Our work greatly accelerates fitting linear mixed models for various genetic study applications. We can now analyze thousands of observations in two minutes rather than waiting half an hour for the results, without compromising the quality of the results,” said Yue.