Table 2 Selected tools and resources for the identification of malignant cells in scRNA-seq data
From: Identification of malignant cells in single-cell transcriptomics data
Resource | Type/readout | Comments | Availability and references |
|---|---|---|---|
InferCNV | Copy number alterations | Arguably the most widely used method for CNA detection in scRNA-seq | |
CopyKAT | Among top performers in recent benchmarks, especially when using only gene expression matrix | ||
Numbat | Exploits allelic imbalance to improve CNA prediction; requires sequencing reads | ||
LISI | Inter-patient heterogeneity | A simple metric of patient mixing | |
scIntegrationMetrics | Implements per-cell-type LISI and additional metrics | ||
scAllele | Single nucleotide alterations | SNA detection tailored for scRNA-seq | |
Monopogen | SNA calling (germline + somatic) leveraging linkage disequilibrium from reference panels | ||
STAR-fusion | Fusion transcripts | Primarily designed for bulk RNA-seq, but can be adapted for single-cell data | |
scFusion | Specific for gene fusion detection at single-cell resolution | ||
UCell | Gene signature scoring | Simple and robust rank-based gene set scoring | |
GSVA | Implements methods for gene set enrichment analysis | ||
scATOMIC | Automated classifier | Integrated pipeline for cell type classification, including malignant vs. normal cells | |
Ikarus | Relies on DEG signatures between normal and malignant cells | ||
scMalignantFinder | Uses logistic regression trained on curated pan‑cancer gene signatures and DEGs | ||
OncoDB | Database | Collates expression profiles for cancer vs. normal tissues | |
3CA | Provides robust transcriptional meta-programs for several cancer types | ||
HPA | Includes scRNA-seq expression profiles for many tissues and cell types |
