Spatial omic technologies have revolutionized tissue analysis by enabling multimodal molecular coprofiling within their native tissue context. Integrating multislice spatial multiomic data offers unprecedented opportunities to reconstruct three-dimensional (3D) tissue landscapes from multimodal molecular perspectives. However, current spatial omic integration methods remain narrowly focused on either vertical (cross-omic) or horizontal (cross-slice) integration, leaving a critical gap for a unified framework that simultaneously addresses both dimensions. Here we present SpatialMOSI, a unified framework for mosaic integration that concurrently resolves cross-modality and cross-section variations. At its core, SpatialMOSI employs a hierarchical graph contrastive learning (HiGCL) strategy that coordinates three integrative objectives: cross-omic alignment and fusion, cross-slice batch correction, and spatial microenvironment preservation. This approach operates on modality-specific latent representations while maintaining feature fidelity through decoding reconstruction. We demonstrate SpatialMOSI's versatility across multiple biological systems, accurately identifying spatially conserved domains, imputing missing omic layers, revealing B cell dynamics in germinal centers, delineating tumor-immune interactions, and reconstructing embryonic developmental trajectories. SpatialMOSI provides a critical computational foundation for constructing integrative 3D molecular atlases from complex multimodal spatial data sets.
Advances in single-cell sequencing technologies greatly enhance our understanding of molecular and cellular features. However, effectively leveraging these data to uncover key biological factors remains a major challenge in integrative analyses across multiomic types and comparative studies across species, particularly livestock species such as pigs and cattle. To address this, we develop AlignCell, a deep learning model designed to learn robust biological features by integrating multisource omic data across platforms, omic types, and species, thereby facilitating the discovery of key factors, such as conserved and species-specific genes in cross-species comparative studies. Across various applications and benchmarking compared with existing tools, AlignCell performs well. Notably, using AlignCell to integrate female gonad data across four species (human, mouse, pig, and cattle), including the bovine single-cell data generated in this study, AlignCell reveals the unexplored species-conserved gene CCT2 in primordial germ cells (PGCs). Additionally, it identifies unexplored species-specific genes PRICKLE4 and CTSV in pig and cattle PGCs, providing important insights for reproductive and developmental research.
Single-cell multiomic technology makes it possible to profile both the transcriptome and epigenome within individual cells. The detection of rare cell states from such data is crucial for exploring novel disease biomarkers with clinical potential. However, existing methods typically focus on the gene and peak counts while ignoring the underlying genomic sequence at accessible sites. Here, we propose ComicGTN, an innovative computational framework that integrates single-cell multiomic data with DNA sequence information via enhanced graph transformer networks to accurately identify rare cell clusters. ComicGTN consistently outperforms nine state-of-the-art methods in identifying rare cells across multiple complex scenarios. In mouse breast cancer data, ComicGTN detects several functionally distinct immune subpopulations. In cerebral cortex of epilepsy patients, it uncovers oligodendrocytes undergoing particular differentiation states. For polycystic kidney disease patient-derived organoids, ComicGTN elaborates pathological pathways in a captured rare glomerular subgroup. Overall, ComicGTN accelerates the localization of disease-associated rare cell populations, facilitating the derivation of clinical insights in development and disease progression.
Meiosis relies on programmed DNA double-strand breaks (DSBs) to initiate recombination between homologous chromosomes. Only a fraction of these breaks mature into crossovers (COs), creating the chiasmata essential for physically linking homologs and ensuring their accurate segregation in meiosis I. Errors in CO number or placement underlie a large fraction of human infertility and aneuploidies, yet how DSBs are designated to become COs and how the mechanisms ensuring COs form on all chromosomes (CO assurance) are not well understood, particularly in systems with a large number of chromosomes. Here, we investigate CO formation in the pantry moth, Plodia interpunctella, a novel invertebrate model system with n = 31 chromosomes. Using a combination of sequencing approaches, Oligopaints FISH, and statistical Weinstein tetrad analysis, we find robust CO assurance and interference in Plodia, with most chromosomes harboring only a single distal CO and few bivalents lacking COs. We employ CUT&RUN, ATAC-seq, and RNA-seq to profile the epigenomic landscape in Plodia testes, revealing that the distal chromosome arms in which COs form are enriched for heterochromatin and devoid of accessible chromatin, suggesting a bias for COs to form in repressed regions in the genome. These studies pave the way for future work diving deeper into the molecular regulation of CO formation in Plodia and other systems, in which the large number of chromosomes may be the key to revealing novel insights about CO regulation.
Many patients do not experience optimal benefits from medical advances because clinical research does not adequately represent them. Although the diversity of biomedical research cohorts is improving, ensuring that individual patients are adequately represented remains challenging. We propose a new approach, GenoSiS, which leverages machine learning-based similarity search to dynamically find patient-matched cohorts across different populations quickly. These cohorts could serve as reference cohorts to improve a range of clinical analyses, including disease risk score calculations and dosage decisions. Although GenoSiS focuses on finding genetic similarity within a biobank, our similarity search architecture can be extended to represent other medically relevant patient characteristics and search other biobanks.
