Ritambhara Singh

The paper introduces GC-MERGE, a graph convolutional model that integrates spatial genomic organization and histone modifications to predict gene expression and interpret regulatory factors.

SCOT uses Gromov-Wasserstein optimal transport to align single-cell multi-omics data without requiring correspondence information for hyperparameter tuning, performing on par with state-of-the-art methods but faster.

The paper improves SCOT for integrating single-cell multi-omics datasets with imbalanced cell-type representation and varying sample sizes, showing better alignment performance on real-world datasets.

The paper integrates single-cell RNA-seq, ATAC-seq, and Hi-C data to study nuclear configuration and gene expression changes during stem cell differentiation and X inactivation, revealing discrete structural changes and delayed chromatin compaction.

TIMEOR identifies temporal regulatory mechanisms from multi-omics data by inferring gene regulatory relationships across time, revealing a link between insulin stimulation and circadian rhythm.

This paper identifies a bias in machine learning models that predict genomic features across cell types when trained on data from the same genomic loci across multiple cell types.

This paper applies and extends MMD-MA to align single-cell transcriptome and chromatin accessibility data from ~2000 cells, demonstrating its scalability and accuracy.

FastSK improves the speed and scalability of gapped k-mer kernel computations for sequence analysis tasks, outperforming existing methods in both speed and accuracy.

MMD-MA embeds cells measured by different single-cell assays into a shared latent space without cell or feature correspondence, recovering true pairings in simulations and a 61-cell expression/methylation co-assay.