Featured | Adrian Salavaty

CelliVerse

Wed, 19 Aug 2026 00:00:00 +0000

Overview

celliverse provides an integrated framework for single-cell RNA-seq analysis with a focus on data-driven clustering, marker discovery, marker assessment and cell type annotation. Its methods use cell-intrinsic expression organisation to support robust characterisation of cellular populations while remaining practical within common single-cell workflows.

The package includes ClustoCell for unsupervised cell clustering and sub-clustering, tools for positive and negative marker discovery, marker purity assessment, dataset-level feature selection, label transfer and annotation workflows. It also provides CelliVerse MarkerDB, a curated marker resource containing positive and negative markers for human and mouse cell types.

Quick links

Feature Explorer Adoption Hub

ClustoCell

Sun, 24 May 2026 00:00:00 +0000

Overview

ClustoCell is a reference-independent framework for resolving cell types and states from single-cell transcriptomes by modelling cellular relationships using cell-intrinsic expression organisation rather than relying on global transcriptional variance or external reference labels.

ClustoCell stratifies expression within each cell and uses this structure to build cell-similarity graphs, identify biologically coherent populations, discover cluster-specific markers and optionally resolve sub-clusters. It is implemented in the CelliVerse R package and is designed to support the recovery of stable cellular communities, rare or transitional states, and malignant versus non-malignant populations.

Beyond clustering, the broader CelliVerse workflow connects ClustoCell results to marker assessment, feature selection and cell type annotation, including annotation with CelliVerse MarkerDB.

Influential R package

Sat, 27 Nov 2021 00:00:00 +0000

Overview

The goal of influential is to help identification of the most influential nodes in a network as well as the classification and ranking of top candidate features. This package contains functions for the classification and ranking of features, reconstruction of networks from adjacency matrices and data frames, analysis of the topology of the network and calculation of centrality measures as well as a novel and powerful influential node ranking. The Experimental data-based Integrative Ranking (ExIR) is a sophisticated model for classification and ranking of the top candidate features based on only the experimental data. The first integrative method, namely the Integrated Value of Influence (IVI), that captures all topological dimensions of the network for the identification of network most influential nodes is also provided as a function. Also, neighborhood connectivity, H-index, local H-index, and collective influence (CI), all of which required centrality measures for the calculation of IVI, are for the first time provided in an R package. Additionally, a function is provided for running SIRIR model, which is the combination of leave-one-out cross validation technique and the conventional SIR model, on a network to unsupervisedly rank the true influence of vertices. Furthermore, some functions have been provided for the assessment of dependence and correlation of two network centrality measures as well as the conditional probability of deviation from their corresponding means in opposite directions.

Interactive Demo

Explore different features of the influential R package in an interactive mode.

Interactive Demo

ExIR

Thu, 27 Feb 2020 00:00:00 +0000

Overview

This model is available in both R and Python versions of the influential package. The tutorial video showcases the function performance in the R environment.

ExIR (experimental data-based integrative ranking) is a data-driven framework designed to identify and prioritize the most biologically relevant genes, proteins, and other molecular features from high-dimensional omics data. By analyzing feature behavior within inferred association networks, ExIR classifies and ranks candidates as drivers, biomarkers, or mediators, helping researchers focus on the features most likely to play important biological roles. Unlike many prioritization approaches, ExIR operates directly on experimental data without requiring external annotations or prior biological knowledge.

ExIR is applicable to virtually all bulk and single-cell omics data types, including large-scale single-cell atlases. To support efficient analysis of very large datasets, it incorporates pseudo-bulking functionality that can process datasets containing millions of cells while substantially reducing computational requirements. ExIR also integrates seamlessly with standard single-cell analysis workflows through native support for Seurat objects.

The framework has been evaluated across multiple transcriptomic and proteomic datasets, consistently demonstrating strong feature-prioritization performance relative to commonly used approaches. In one application, ExIR identified candidate regulators associated with disease progression in RNA-seq data from a zebrafish model of mucopolysaccharidosis IIIA. Together, these capabilities make ExIR a versatile and scalable solution for extracting biologically meaningful insights from modern omics datasets and guiding downstream experimental investigation.

Interactive Demo

IVI

Thu, 27 Feb 2020 00:00:00 +0000

Overview

Biological systems are composed of highly complex networks, and decoding the functional significance of individual network components is critical for understanding healthy and diseased states. Several algorithms have been designed to identify the most influential regulatory points within a network. However, current methods do not address all the topological dimensions of a network or correct for inherent positional biases, which limits their applicability. To overcome this computational deficit, we undertook a statistical assessment of 200 real-world and simulated networks to decipher associations between centrality measures and developed an algorithm termed Integrated Value of Influence (IVI), which integrates the most important and commonly used network centrality measures in an unbiased way. When compared against 12 other contemporary influential node identification methods on ten different networks, the IVI algorithm outperformed all other assessed methods. Using this versatile method, network researchers can now identify the most influential network nodes. This function is available in both R and Python versions of the influential package. The tutorial video showcases the function performance in the R environment. You can also visualize a network based on its IVI values using the Centrality-based Network Visualization function.

Interactive Demo