Bokan Bao

The paper investigates the phenotypic and transcriptional effects of glycoengineering in CHO cells, identifying distinct clusters and molecular mechanisms associated with different glycosyltransferase modifications.

The paper trains a hybrid neural network to predict N-glycosylation events and structures from protein sequences using glycosite-flanking data, demonstrating its ability to predict glycan composition and abundance.

The paper identifies a relationship between protein structure and glycan structure, introducing a glycoimpact matrix to predict glycosylation patterns and validate it through disease variants and engineered antibodies.

GlyCompareCT preprocesses glycomics data to address sparsity and non-independence, improving statistical power for downstream analysis.

The paper constructs an HMO biosynthesis network using multi-omics integration to predict glycosyltransferases and propose candidate genes for HMO synthesis pathways.

This paper improves ASD outcome prediction from gene expression using transfer learning, achieving a 7.8% AUROC improvement over a naive model.

This paper extends Linear Code to represent reaction rules in glycobiology, aiming to improve consistency and interoperability in glycan synthesis models.

The paper introduces GlyCompare, a glycomic analysis method that addresses sparsity and non-independence by leveraging biosynthetic pathways, revealing biological insights in glycoprofiles and human milk oligosaccharides.

HiCluster improves clustering of low coverage single-cell Hi-C data by imputing using linear convolution and random walk, revealing TAD-like structures consistent with bulk data.