Pavlo Lutsik

The paper develops a hierarchical classification framework combining epigenetic and genetic data with machine learning to improve diagnosis and risk stratification in hematologic malignancies.

This study shows that loss of the Y chromosome in lung adenocarcinoma causes epigenetic and transcriptomic changes that increase lineage plasticity and poor clinical outcomes.

This study identifies GLYATL1 as a mediator of endocrine resistance in luminal breast cancer by promoting metabolic and epigenetic changes that counteract estrogen deprivation.

The paper investigates DNA methylation differences in prostate cancer across African, European, and Asian ancestries, finding ancestry-driven methylation patterns linked to disease aggressiveness in African tumors.

The paper benchmarks DNA methylation sequencing workflows against an experimental gold-standard, identifying superior workflows and developing a sustainable benchmarking platform.

Pyjacker detects enhancer hijacking in AML, identifying MNX1 activation via 7q deletion as a novel oncogenic mechanism.

MethylBERT is a Transformer-based model that identifies tumour-derived reads and estimates tumour purity from bulk samples, outperforming existing methods in accuracy.

This study shows that inhibiting the DNMT/G9a complex targets specific epigenetic vulnerabilities in AML, inducing an IFN response and improving therapeutic outcomes.

The paper introduces epiCHAOS, a metric to quantify cell-to-cell epigenomic heterogeneity in single-cell data and shows its utility in assessing stemness and plasticity.

The paper identifies del(5q) AML as an epigenetically distinct subgroup driven by KDM3B haploinsufficiency, linked to DNA hypermethylation and H3K9me1/2 dysregulation.

This study shows that SRSF2 mutations cause DNA damage and cell cycle arrest, highlighting its role in transcription regulation during cell division.

scMaui is a deep learning framework that integrates single-cell multiomics data, handles batch effects, and addresses missing data, showing improved performance in downstream analyses.

Figeno generates publication-quality genomic figures with support for multi-region views, long reads, and epigenomic data.

This study shows that Dnmt3a1, but not Dnmt3a2, is required for hippocampus-dependent memory through its regulation of Nrp1.

The paper develops a computational framework to identify regulatory heterogeneity in leiomyosarcoma by analyzing gene regulatory networks and validates findings in independent datasets.

The paper develops a computational framework for analyzing DNA methylation data in mouse hematopoiesis, identifying novel cis-regulatory elements associated with gene expression during differentiation.

This study benchmarks six sequencing-based and two array-based cell-type deconvolution methods for bulk DNA methylomes, identifying the best-performing method and factors affecting performance.

AmpliconDesign provides a web server for designing high-throughput targeted DNA methylation assays, improving primer design and analysis workflows for epigenetic studies.

This paper introduces a reference-free deconvolution protocol for DNA methylomes to identify hidden cell populations in complex samples, applied to lung cancer data from TCGA.

This study evaluates DNA methylation deconvolution methods for cell-type proportion inference, identifies critical steps for accuracy, and develops a benchmark pipeline.

The paper identifies DNA methylation differences in monozygotic twins discordant for multiple sclerosis, highlighting treatment effects and genetic confounders.