Ivan G. Costa

AtlasOT improves multimodal integration of disease atlases by addressing unpaired data constraints through a novel optimal transport framework.

This paper identifies PDE3A-SLFN12 molecular glues as a novel therapeutic target for KIT D816V-driven mast cell malignancies, showing their efficacy in killing multiple KIT D816V cell types and synergy with tyrosine kinase inhibitors.

This study investigates the relationship between age-related epigenetic changes in DNA methylation, chromatin accessibility, and histone modifications across human aging using integrated WGBS and ATAC-seq data.

IntraTalker+CrossTalkeR models intracellular and intercellular signaling in single-cell data to uncover cell-state changes and novel signaling roles.

This study uses multimodal single-cell analysis to identify transcription factor networks involved in T-cell aging.

GraphHDBSCAN* performs hierarchical clustering on scRNA-seq data, recovering biologically meaningful hierarchies and improving flat partition quality.

This study identifies Spp1⁺ macrophages as key hubs coordinating fibrotic remodeling in myeloproliferative neoplasms using NicheSphere, a novel computational method for cell communication analysis.

This study identifies polyploid cardiomyocytes as disease-specific transcriptional states in the mammalian heart and suggests targeting them for heart failure treatment.

The paper shows that Wnt-dependent reprogramming of bone marrow niches drives fibrosis by activating osteolineage cells and altering hematopoietic support.

PILOT-GM-VAE analyzes single cell disease atlases to perform sample-level clustering, trajectory inference, and batch correction using Gaussian Mixture Variational Autoencoders with optimal transport.

This paper benchmarks clustering algorithms for scRNA-seq, showing that hyperparameter choices and graph construction significantly affect performance, with no practical difference between Leiden and Louvain algorithms.

This study shows that CREB knockout in Foxp3+ Tregs increases IL-10 and ST-2 expression, enhancing their suppressive function and preventing gut inflammation.

This study investigates the postnatal maturation of the small intestinal epithelium, identifying developmental regulators as key drivers over microbiota effects.

scACCorDiON clusters patient-specific cell-cell communication graphs using optimal transport, improving disease status prediction in scRNA-seq data.

The paper introduces a weighted 2D-kernel density estimation method for epigenetic age prediction, achieving similar accuracy to existing clocks and providing variation scores linked to mortality risk.

This study identifies JAK2V617F zygosity-specific vascular changes in MPN and shows IFNα inhibits EndMT, offering a novel antifibrotic mechanism.

PHLOWER infers complex differentiation trees from multi-modal single-cell data using Hodge Laplacian decomposition, showing effectiveness in kidney organoid data.

SIGURD quantifies clones in scRNA-seq data by integrating somatic and mitochondrial variants and enables functional analysis of detected clones.

The paper identifies aberrant DNA methylation in cancer-associated fibroblasts across multiple cancer types and links these changes to patient prognosis.

PILOT uses optimal transport to analyze single-cell and pathomics data, detecting disease-associated samples and non-linear changes in cell populations and tissue structures.

CimpleG identifies small CpG methylation signatures for cell type classification and deconvolution, performing efficiently with single CpG sites per cell type.

The paper shows ephrinA5 reduces cell motility by decreasing Snhg15 expression, which lowers DNMT1 binding to the Ncam1 promoter, reducing methylation and increasing Ncam1 expression.

RGT is a Python package for integrative analysis of regulatory genomics data, offering tools for peak calling, binding site prediction, and regulatory factor association.

scMEGA infers gene regulatory networks from single-cell multi-omics data, demonstrating its application in studying myofibroblast activation in human myocardial infarction.

The paper identifies HIF1α-AS1 as a DNA:DNA:RNA triplex-forming lncRNA that interacts with the HUSH complex to repress gene expression and modulates angiogenesis in pulmonary hypertension.

This study generates a high-resolution spatial multi-omic map of human myocardial infarction by integrating single-cell gene expression, chromatin accessibility, and spatial transcriptomic data to reveal cellular zones of injury, repair, and remodeling.

The paper compares 15 resources and 6 methods for inferring cell-cell communication from single-cell RNA data, finding significant variation in predicted interactions and biological interpretations.

This study uses human pluripotent stem cell-derived kidney organoids to model nephrotic syndrome, showing they can replicate podocyte injury and respond to genetic correction.

MOJITOO integrates multimodal single cell data using canonical correlation analysis, outperforming existing methods in computational efficiency and clustering.

scOpen improves downstream analysis of sparse scATAC-seq data by imputing open chromatin regions, revealing regulatory changes in kidney fibrosis.