Abstract
We present a case study applying hierarchical Bayesian estimation on high-throughput protein melting-point data measured across the tree of life. We show that the model is able to impute reasonable melting temperatures even in the face of unreasonably noisy data. Additionally, we demonstrate how to use the variance in melting-temperature posterior-distribution estimates to enable principled decision-making in common high-throughput measurement tasks, and contrast the decision-making workflow against simple maximum-likelihood curve-fitting. We conclude with a discussion of the relative merits of each workflow.
Highlights
Workflow with Hierarchical BayesianHigh-throughput measurements are a staple of biological measurements
A wealth of literature exists for the statistical analysis of high-throughput measurement data [1,2,3]
Other examples include its application in functional magnetic resonance imaging [7], survival analysis [8], genomic population analysis [9] and infectious agent risk analysis [10]
Summary
Workflow with Hierarchical BayesianHigh-throughput measurements are a staple of biological measurements. A wealth of literature exists for the statistical analysis of high-throughput measurement data [1,2,3]. To the best of our knowledge, the application of hierarchical Bayesian models in high-throughput assay measurements is not widespread, with only a countably small number of papers leveraging hierarchical Bayesian methods [4,5]. The advantages of hierarchical models for estimation are well-known. Players with fewer replicate observations of their fielding statistics had estimates shrunk closer to the population mean, though as more replicate measurements are obtained, the regularization effect diminishes. The main advantage of hierarchical models is that they allow for pooling of information between samples in a principled fashion, while still allowing for individual differences to show up, provided there is sufficient evidence for these differences
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
