Olga V. Kalinina

TopFor automates the generation of AMBER-compatible parameters for noncanonical amino acids in peptides, enabling efficient MD simulations and integration into peptide discovery workflows.

This study identifies information leakage in enzyme-substrate prediction models and datasets, showing performance drops when leakage is removed.

The paper presents 154 new Myxococcota genomes, expands taxonomy, and identifies novel natural products, highlighting their biosynthetic potential for drug discovery.

ALPAR automates AMR data analysis from genomic data processing to ML model training and GWAS, offering a reproducible pipeline for single-reference AMR studies.

The paper evaluates zero-shot models' ability to predict epistatic effects on protein properties and finds they fail to predict strongly epistatic mutation combinations.

The paper identifies compact, data-driven codes for adenylation domain substrate specificity in NRPSs, outperforming traditional codes and revealing new positions critical for specificity.

The paper evaluates XAI methods on GNNs for DTI prediction, finding that consensus attributions highlight atoms in binding pockets relevant to drug-target interactions.

This paper uses generative AI to design functional thiolation domains for reprogramming NRPSs, improving product formation and stability.

LoReMINE is an automated pipeline for microbial genome mining that integrates assembly, classification, and biosynthetic gene cluster prediction from long-read data.

StructGuy predicts functional effects of genetic variants using a data leakage-free supervised approach, achieving competitive performance against zero-shot methods.

This paper introduces EMMA, a transformer-based multi-task learning framework for predicting enzyme-small molecule interactions, showing strong performance across different data-splitting regimes.

This study investigates alternative splicing and miRNA regulation in a zebrafish FSGS model, identifying key splicing events and miRNAs associated with disease progression.

The paper investigates alternative splicing in mechanically stretched podocytes, identifying splicing events in Shroom3 and Myl6 that may contribute to hypertensive nephropathy.

EpiPAMPAS detects intra-protein epistasis by analyzing multiple sequence alignments and finds interactions with low p-values in simulated and real viral protein data.

DataSAIL splits biological datasets to minimize information leakage by reducing similarity between splits, improving model generalization assessment.

The paper introduces a phylogeny-related parallelism score to improve machine learning prediction of antibiotic resistance in M. tuberculosis by incorporating evolutionary relationships.

PanPA generates and aligns panproteome graphs to improve alignment accuracy in prokaryotic genomes by leveraging amino acid similarity over DNA sequence diversity.

GlyLES converts IUPAC-condensed glycan notation to SMILES using a grammar-based approach, enabling atomic-level representation and applications in molecular modeling and machine learning.

MetaProFi is a Bloom filter-based tool that stores and queries sequence data, enabling efficient and accurate identification of functionally relevant genetic variants by supporting amino acid and translated nucleotide sequences.

The paper presents an extended catalogue of tandem alternative splice sites in human tissues, analyzing their expression and regulation.

This paper introduces a method to identify biologically meaningful structural motifs in protein 3D structures using frequent subgraph mining, validated on known motifs and novel discoveries.