Souvik Seal

RKMR identifies predictive marker panels in spatial omics data using a scalable framework that integrates kernel modeling and spatial dependence.

FusedFCR identifies marker genes associated with developmental transitions in time-course scRNA-seq data by combining lasso and fusion penalties for interpretable gene selection.

MSGPCA improves spatial omics analysis by decomposing molecular variation into shared and slice-specific factors, enhancing the detection of tissue structure and microenvironmental variation.

The paper compares T-cell counts and spatial patterns in colorectal adenomas and sessile serrated lesions, finding differences in T-cell infiltration and clustering by lesion type.

SpaceBF identifies spatially varying co-expression patterns in spatial omics data using a Bayesian fused modeling framework, outperforming existing methods in specificity and power.

CBKMR improves marker detection in omics data by capturing nonlinear effects and gene interactions, outperforming existing methods in accuracy and uncertainty quantification.

The paper introduces a Bayesian negative binomial regression model with group-structured priors for spatially correlated count data, applied to cancer registry and spatial omics datasets.

cytoKernel assesses differential expression in single-cell data by using kernel embeddings to detect both aggregate and subtle variations, outperforming existing methods.

SpaceANOVA analyzes spatial co-occurrence of cell types in multiplex imaging data across tissue or disease groups using point process and functional ANOVA, revealing novel insights in colorectal cancer precursor lesions.

SMASH identifies spatially variable genes in spatial transcriptomics data with improved statistical power and computational efficiency.

The paper introduces FunSpace, a method using spatial entropy and functional principal component analysis to assess spatial heterogeneity in tumor microenvironments and its impact on survival.

This paper introduces RCFGL, a fast method for joint estimation of condition-specific gene co-expression networks, showing improved computational efficiency and performance over existing methods.

The paper introduces MIAMI, a method using mutual information to analyze protein co-expression in cancer datasets, finding associations between marker co-expression and clinical outcomes.

The paper estimates SNP heritability in multi-ethnic data while accounting for population substructure using a modified method of moments approach.

DenVar analyzes multiplex imaging data to assess associations between functional marker densities and clinical outcomes without thresholding.

This paper introduces PredLMM, a method for rapidly estimating SNP heritability in large cohorts by approximating linear mixed models using predictive process techniques.