Joakim Jaldén

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

The paper introduces a Brownian motion data augmentation method to improve neural network performance on nanopore sensors, achieving higher accuracy with a new network architecture.

Lokatt improves nanopore basecalling by combining an explicit duration HMM with a residual LSTM, achieving high accuracy and coverage on E. coli data.

The paper introduces a beam search decoder that improves decoding speed for single-molecule peptide sequencing data with slightly higher accuracy than existing methods.