Zhengyang Guo

This study identifies Janusein as a microtubule-associated protein that cooperates with EB proteins to enhance microtubule plus-end catastrophe.

The paper develops a framework to rerank protein-protein interactions using structural and biological evidence, improving prediction accuracy over structural-only methods.

This paper develops a hierarchical sequence-based framework to infer distributed protein organization, identifies novel ciliary proteins, and enables synthetic protein design.

TaoChongBao is a database of C. elegans missense variants linked to human pathogenic sites, offering expanded mutation coverage and tools for variant interpretation and comparative genomics.

This study identifies GPCR-Gαq signaling as a mechanism of adaptive resistance to MEK inhibition in BRAF-mutant melanoma, showing that targeting this pathway can restore drug sensitivity.

The paper systematically assesses the functional impact of all TUBA1A missense variants using experimental and computational methods, improving clinical variant classification.

This study identifies FAP53 as a key factor in B-tubule assembly during microtubule doublet formation and shows its role in ciliogenesis across species.

SynSeg generates synthetic datasets to train a U-net model for subcellular segmentation, outperforming traditional methods in vesicle and cytoskeletal filament detection.

This study shows that phosphorylation of the elbow region in kinesin OSM-3 regulates its autoinhibition, with phosphorylation mutants affecting cilia entry and motility.

The paper develops an AI-driven method to label tubulin isotypes in vivo across species, enabling precise visualization of microtubule dynamics and pathogenic variants without disrupting cellular processes.

The paper introduces a cost-effective WGS method for model organisms to validate missense variant predictions.

The paper develops a machine learning-based EMS mutagenesis probability map to identify causal mutations in C. elegans without generating recombinant inbred lines.

ADAMS aligns distance matrices using SIFT to improve protein structure comparison, matching Foldseek's performance while better handling flexible structures.