Ritambhara Singh

scWeave translates between gene expression and chromatin structure at single-cell resolution using deep learning, improving cross-modality inference across multiple datasets.

DRIFT improves spatial transcriptomics analysis by integrating spatial context into foundation models using diffusion on spatial graphs, outperforming specialized methods.

The paper develops a deep learning model called PLASMA to predict early recurrence of lung adenocarcinoma using clinical, mRNA expression, and mutation data, achieving high AUROC scores in validation sets.

ACE-OF-Clust improves single-cell clustering by aligning results across methods and identifying informative features in scRNA-seq and spatial transcriptomics data.

The paper introduces TimeFlies, an snRNA-seq aging clock for Drosophila heads, identifies key lncRNAs as biomarkers, and shows sex-specific aging pathways.

WINDEX improves detection of positive selective sweeps by integrating site- and window-based statistics in population genetic data, showing improved localization and genome-wide estimates of selective pressures.

The paper introduces ctOTVelo, a method for inferring cell type-specific gene regulatory networks from single-cell transcriptomics data, achieving state-of-the-art performance.

This study compares transcriptomic profiles of embryonic stem cells across mammals, identifying conserved and species-specific pluripotency mechanisms and markers.

SCOT+ aligns single-cell multi-omics data across datasets using optimal transport without requiring correspondence, improving downstream analysis accuracy.

The paper identifies sex-biased gene expression and age-related patterns in painted turtles, linking female-biased longevity to molecular differences without sex chromosomes.

The paper introduces eccDNAMamba, a model for ultra-long eccDNA sequences, showing improved performance in cancer discrimination and copy-number prediction.

The paper improves gene regulatory network inference by integrating cell cycle positions into oscillatory network analysis, showing better performance for early progenitor cells.

STING improves spatial transcriptomics clustering by integrating gene-gene and spatial relations through nested graph neural networks, achieving better clustering performance across multiple datasets.

HypoAD integrates MRI and snRNA-seq to analyze hypothalamic changes in aging and Alzheimer's disease, revealing volume and gene expression alterations in key subregions and cell types.

This study identifies sex-biased gene expression, alternative splicing, and isoform dynamics in bovine embryos, linking male-biased metabolism to faster development and female-biased X-linked gene expression to sex-specific regulation.

LUNAR predicts glioma recurrence by integrating clinical, genomic, and mRNA expression data from grade II-IV glioma patients.

OTVelo estimates gene velocities using optimal transport to infer dynamic gene regulatory networks from time-stamped single-cell gene expression data.

This study identifies molecular signatures of ancestral cell types in mammalian olfactory cortex neurons by comparing transcriptomes across species and cortical layers.

scContrast is a contrastive learning method for embedding scRNA-seq data into a universal space, outperforming existing methods in integrating data from different sequencing protocols.

scMultiNODE integrates gene expression and chromatin accessibility data across timepoints to recover developmental dynamics and annotations for weak modalities.

GrapHiC improves Hi-C read imputation by integrating Hi-C and ChIP-seq data in a graph framework, enhancing performance on sparse and cross-cell-type inputs.

scGrapHiC uses deep learning to predict cell-type-specific scHi-C contact maps from scRNA-seq data, improving recovery of Topologically Associating Domains by 23.2% over baselines.

scNODE predicts single-cell gene expression at unobserved timepoints using a deep learning model combining VAE and ODEs, improving trajectory inference and perturbation analysis.

This paper introduces a deep learning model to predict A/B compartments from histone modifications, achieving high accuracy across cell types and showing the importance of specific histone marks.

This paper evaluates deep-learning methods for improving Hi-C resolution using experimentally generated datasets and finds they perform poorly on sparse data, with retraining offering limited improvement.

The paper develops histone mark-based age prediction models across tissues, showing comparable accuracy to DNA methylation predictors and identifying key developmental pathways involved in aging.

This study evaluates gene co-expression estimation methods on simulated and experimental single-cell datasets, finding high false discovery rates and limited improvement from preprocessing.

This paper introduces a co-optimal transport method for aligning cells and genomic features in unpaired single-cell multi-omic datasets, improving integration with sparse or dense feature correspondences.

AltumAge is a deep learning-based DNA methylation clock that improves age prediction accuracy and interpretability across tissues and ages.