Maria Jesus Martin

PDBe-SIFTS improves residue-level sequence-structure mapping with faster search, better scoring, and accurate alignment, comparable to expert curation.

The paper identifies artifacts in bacterial proteomes by analyzing cluster participation and protein length, finding that short proteins are often errors due to frame-shifts or termination codons.

The paper creates a high-confidence human methylation atlas and develops a deep learning model for methylated peptide detection.

This work analyzes CAFA5 submissions, showing improved accuracy in protein function prediction and highlights the role of protein language models and 3D structures in performance gains.

The paper introduces a method to select Reference Proteomes for UniProt, increasing coverage and reducing redundancy in the knowledgebase.

The paper identifies 25% of 7,264 non-canonical open reading frames (ncORFs) as producing translated gene products, supported by over 3,000 peptides from 3.8 billion mass spectra.

The paper maps the functional importance of site-specific ubiquitination across the human proteome by analyzing conservation, perturbation, and machine learning scores.

The paper develops a pan-gene catalogue for Asian cultivated rice by creating a reference panel of 16 cultivars and identifying core and cultivar-specific genes.

The paper predicts the mechanistic impacts of human protein missense variants on stability, protein-protein interactions, and small-molecule binding using AlphaFold2 structures.

The paper improves canonical protein selection for reference proteomes by identifying more accurate representatives among orthologous sequences.

This paper performs a meta-analysis of 11 phosphoproteomics datasets to identify phosphorylated sites in Plasmodium falciparum and related species, revealing sequence motifs, conservation patterns, and structural context of phosphosites.

This study analyzes rice phosphoproteomics data to identify phosphosite variations across the rice pan-genome and their functional implications.

The paper reannotates ligand binding sites in UniProtKB using ChEBI identifiers and improves search facilities for these annotations.

This paper outlines a community-driven initiative to create a consolidated catalog of human Ribo-seq ORFs, addressing the lack of standardized annotation for these features.

The paper estimates the false discovery rate of phosphosites in public databases and predicts a more realistic count of true phosphosites in the human proteome.

DRUIDom predicts compound-target interactions using protein domain modularity, identifying 27,032 high-confidence associations and validating them experimentally on LIM kinases.

The paper curates a comprehensive catalogue of human DNA-binding transcription factors using Gene Ontology definitions and provides access to it via an online resource.

The paper introduces a method to independently estimate the false localisation rate in phosphoproteomics using decoy amino acids, showing that alanine and leucine provide reliable results.

This paper describes a workflow to improve glycan data interoperability between ChEBI, PubChem, and GlyGen by integrating and prioritizing glycan annotations.

The paper introduces CROssBAR, a resource that integrates biomedical data and uses deep learning to predict entity relations, displaying results in knowledge graphs for systems-level analysis of diseases like COVID-19.

DEEPScreen predicts drug-target interactions using convolutional neural networks with 2-D compound representations, achieving better or comparable performance to existing methods.

The paper maps UniProtKB protein annotations to the human genome and demonstrates their utility in interpreting genetic variants.

The paper outlines ten lessons for designing, provisioning, and reusing persistent identifiers to improve data integration in life sciences.

BioJS-HGV Viewer provides an interactive tool to visualize genetic variants from multiple sources, enabling detailed exploration of variant sites and annotations.