Biology subjects

Mann-Krzisnik, D.

Publications and source records attributed to Mann-Krzisnik, D..

3 recordsLinked to original sources

bioRxiv2026

SpatialJEPA: JEPA-inspired graph-context distillation for spatially aware multiomics integration

Computational frameworks for integrating spatial genomics modalities extend cell-based representation learning across molecular layers, but many paired RNA-ATAC datasets are dissociated and lack spatial coordinates. We introduce SpatialJEPA, a JEPA-inspired teacher-student framework for transferring spatial context from spatial multiomics data to non-spatial multiome data. In contrast to patch- or feature-masking objectives, SpatialJEPA masks spatial context by replacing the teachers spatial neighborhood graph with a self-only identity graph during student training, making the spatial sample appear dissociated to the student. The student learns to match teacher embeddings from this graph-context-restricted view and can therefore be applied to dissociated RNA-ATAC data at inference time. In mouse brain multiomics, the resulting representation supports source-target alignment, recovers spatially organized transcriptomic and chromatin-accessibility programs, and shows concordance with ligand-receptor pathway structure compared with non-spatial references.

molecular biology↗

bioRxiv2025

ECLARE: multi-teacher contrastive learning via ensemble distillation for diagonal integration of single-cell multi-omic data

Integrating multimodal single-cell data such as scRNA-seq and scATAC-seq is key for decoding gene regulatory networks. Still, integration remains challenging due to issues related to feature harmonization and limited quantity of paired data. To address these challenges, we introduce ECLARE, a novel framework combining multi-teacher ensemble knowledge distillation with contrastive learning for integrating unpaired single-cell multi-omic data. Briefly, ECLARE trains teacher models on paired datasets to guide a student model for aligning unpaired data, leveraging a refined contrastive objective and optimal-transport-based loss for precise cross-modality alignment. In computational benchmarking, experiments demonstrate ECLAREs competitive performance in cell pairing accuracy, multimodal integration and biological structure preservation, indicating that multi-teacher knowledge distillation provides an effective means to improve a diagonal integration model beyond its zero-shot capabilities. In biological case studies, we demonstrate ECLAREs applicability using unpaired snRNA-seq and snATAC-seq datasets in major depressive disorder (MDD). Firstly, our results revealed transcription factors and target gene combinations differentially regulated in depression with sex- and cell-type specificity. These findings further reveal gene regulatory interactions in excitatory neurons that are highly relevant to MDD neuropathology, such as those involving EGR1, SOX2, and NR3C1. Secondly, we show that ECLARE can learn continuous data manifolds useful for deciphering longitudinal biological processes in neurodevelopment and disease, revealing altered neurodevelopmental programs as potential regulators of depression in females, strongly associated with EGR1 target genes. Altogether, we propose ECLARE as a robust solution for diagonal integration of unpaired multimodal single-cell data that enables the study of altered gene regulation in disease.

bioinformatics↗

bioRxiv2023

Single-cell multi-omic topic embedding reveals cell-type-specific and COVID-19 severity-related immune signatures

The advent of single-cell multi-omics sequencing technology makes it possible for re-searchers to leverage multiple modalities for individual cells and explore cell heterogeneity. However, the high dimensional, discrete, and sparse nature of the data make the downstream analysis particularly challenging. Most of the existing computational methods for single-cell data analysis are either limited to single modality or lack flexibility and interpretability. In this study, we propose an interpretable deep learning method called multi-omic embedded topic model (moETM) to effectively perform integrative analysis of high-dimensional single-cell multimodal data. moETM integrates multiple omics data via a product-of-experts in the encoder for efficient variational inference and then employs multiple linear decoders to learn the multi-omic signatures of the gene regulatory programs. Through comprehensive experiments on public single-cell transcriptome and chromatin accessibility data (i.e., scRNA+scATAC), as well as scRNA and proteomic data (i.e., CITE-seq), moETM demonstrates superior performance compared with six state-of-the-art single-cell data analysis methods on seven publicly available datasets. By applying moETM to the scRNA+scATAC data in human bone marrow mononuclear cells (BMMCs), we identified sequence motifs corresponding to the transcription factors that regulate immune gene signatures. Applying moETM analysis to CITE-seq data from the COVID-19 patients revealed not only known immune cell-type-specific signatures but also composite multi-omic biomarkers of critical conditions due to COVID-19, thus providing insights from both biological and clinical perspectives.

bioinformatics↗