Natarajan Kannan

CLASPP: A unified model for predicting post-translational modifications

CLASPP is a unified deep learning model that predicts multiple PTM types from protein sequences, addressing data imbalance and improving prediction performance across organisms.

Mechanistic basis of teichoic acid transport by a gatekeeper flippase

The paper determines the mechanism of teichoic acid transport by the flippase TacF in Streptococcus pneumoniae using cryo-EM and computational methods.

Identification and classification of ion-channels across the tree of life provide functional insights into understudied CALHM channels

The paper identifies and classifies ion channels across species, revealing functional insights into understudied CALHM channels through sequence, structure, and functional data integration.

Glydentify: An explainable deep learning platform for glycosyltransferase donor substrate prediction

Glydentify predicts donor substrates for glycosyltransferases using deep learning, achieving high accuracy and providing explainable insights into enzyme-substrate relationships.

Task Splitting and Prompt Engineering for Cypher Query Generation in Domain-Specific Knowledge Graphs

This paper introduces Prompt2Cypher, a method that uses task splitting and prompt engineering to improve Cypher query generation for domain-specific knowledge graphs, showing better accuracy in biological datasets.

Atlas of the Bacterial Serine-Threonine Kinases expands the functional diversity of the kinome

The paper classifies bacterial STK sequences into families and identifies distinguishing features, providing insights for drug discovery.

Dark kinase annotation, mining and visualization using the Protein Kinase Ontology

The paper extends the Protein Kinase Ontology (ProKinO) to uncover new insights about dark kinases, identifying PAK5 as a frequently mutated kinase in cancers and highlighting its potential in drug discovery.

A comprehensive in vivo screen of yeast farnesyltransferase activity reveals broad reactivity across a majority of CXXX sequences

The paper evaluates all possible CXXX sequences as farnesyltransferase targets in yeast using a reporter system, revealing expanded specificity beyond the canonical CaaX motif.

An explainable unsupervised framework for alignment-free protein classification using sequence embeddings

This paper introduces an alignment-free protein classification framework using sequence embeddings to infer evolutionary relationships and functional motifs across diverse protein superfamilies.

Mapping the glycosyltransferase fold landscape using deep learning

This paper introduces a deep learning model to predict glycosyltransferase folds using secondary structure features, identifying novel fold families and improving fold classification.

Quantitative Structure-Mutation-Activity Relationship Tests (QSMART) Model for Protein Kinase Inhibitor Response Prediction

The paper introduces QSMART, a model that predicts protein kinase inhibitor response by combining statistical and neural network methods, achieving high accuracy and feature explainability.

Exploring the understudied human kinome for research and therapeutic opportunities

This paper refines the definition of the human kinome, identifies understudied kinases, and highlights their potential as drug targets through differential expression, essentiality, and mutation data.

Deep evolutionary analysis reveals the design principles of fold A glycosyltransferases

The paper uses evolutionary analysis of GT-A fold sequences to identify core components and hypervariable loops contributing to catalytic diversity, and develops a machine learning classifier for donor specificity prediction.

Repurposing covalent EGFR/HER2 inhibitors for on-target degradation of human Tribbles 2 (TRIB2) pseudokinase

The paper identifies EGFR/HER2 inhibitors that destabilize and degrade TRIB2 pseudokinase in cancer cells, suggesting potential for TRIB2-targeted therapies.

Extensive non-canonical phosphorylation in human cells revealed using strong-anion exchange-mediated phosphoproteomics

This study identifies extensive non-canonical phosphorylation in human cells using a new phosphoproteomics method, revealing over 1000 phospholysine sites and 310 phosphohistidine sites.