Matthew Beauregard Smith

This paper introduces an expectation maximization framework for inferring protein abundances from fluorosequencing data, showing significant improvements in accuracy and scalability.

The paper develops and compares two methods to estimate error rates in single molecule protein sequencing experiments, finding that one method outperforms the other in most criteria.

The paper introduces whatprot, a hybrid kNN-HMM classifier for amino acid sequence assignment from fluorosequencing data, improving precision and recall over individual methods.