Sewoong Oh

Improvements to Casanovo, a deep learning de novo peptidesequencer

The paper improves Casanovo, a deep learning de novo peptidesequencer, by enhancing interpretability, generalizing its use in database search, speeding up performance, and adding tools for adoption and result interpretation.

Accounting for digestion enzyme bias in Casanovo

The paper modifies the Casanovo model to account for digestion enzyme bias but finds no significant improvement in sequencing accuracy.

Sequence-to-sequence translation from mass spectra to peptides with a transformer model

Casanovo translates tandem mass spectra into peptide sequences using a transformer model, outperforming existing methods and improving analysis in immunopeptidomics and metaproteomics.

A learned score function improves the power of mass spectrometry database search

The paper re-purposes a machine learning model for de novo peptide sequencing as a score function for database search, improving detection power over existing methods.

A transformer model for de novo sequencing of data-independent acquisition mass spectrometry data

This paper introduces a transformer-based model called Cascadia for de novo sequencing of data-independent acquisition mass spectrometry data, achieving state-of-the-art performance.

Evaluating proteomics imputation methods with improved criteria

This paper evaluates proteomics imputation methods using downstream-centric criteria and finds that MissForest performs best, though imputation may not improve differential expression detection.

Reducing peptide sequence bias in quantitative mass spectrometry data with machine learning

This paper introduces a deep learning model to reduce sequence-specific biases in quantitative mass spectrometry data, improving correlation with mRNA expression.

Reconstruction of visual images from mouse retinal ganglion cell spiking activity using convolutional neural networks

This paper trains convolutional neural networks on retinal ganglion cell firing data to reconstruct images, showing that subpopulations and ON-OFF cells contribute to image reconstruction.

De novo mass spectrometry peptide sequencing with a transformer model

Casanovo, a transformer encoder-decoder, translates unbinned tandem mass spectrum peaks into peptide sequences, beating DeepNovo and PointNovo on peptide-level precision across nine held-out species while running faster than both.