Table 1 Deconvolution algorithms developed for bulk transcriptomics with sc/snRNA-seq reference datasets. The table includes the name and reference (column 1) along with the year published (column 2) and a description (column 3) of the algorithm. The primary tissues used in the publication associated with the algorithm are also provided (column 4)
Algorithm | Citation | Year | Description | Primary publication tissues |
|---|---|---|---|---|
BayesPrism | [29] | 2022 | Bayesian approach, joint posterior inference and posterior summing over cell states, explicit cell type expression modeling | Blood, multiple cancer types |
Coex | [30] | 2022 | Marker co-expression networks and network module attribution | Brain |
MuSiC2 | [22] | 2021 | Differential marker weighting and filtering on condition-specific differential expression | Pancreas and retina |
SCDC | [31] | 2021 | Ensemble framework to integrate references across sources | Pancreas and mammary gland |
Bisque | [32] | 2020 | Gene-specific transformations to address assay-specific biases | Adipose and brain |
DWLS | [33] | 2019 | Dampened weighted least squares, rare cell type detection | Blood, tumor/melanoma (human); kidney, lung, liver, small intestine (mouse) |
MuSiC | [28] | 2019 | Differential marker weighting to address marker expression confounding | Pancreas and kidney |
dtangle | [34] | 2019 | Marker selection with linear mixed modeling | Blood, breast, brain, liver, lung, muscle, cancer |
ABIS | [35] | 2019 | Absolute deconvolution with cell scale factors on TPM-normalized marker expression | Blood and immune cells |
quanTIseq | [36] | 2019 | Non-negative regression with cell factor scaling and unknown cell type estimation | Blood and tumor |
Fardeep | [37] | 2019 | Machine learning with adaptive trimmed least squares | Immune cells [38], tumor cells (GSM269529) |
BrainInABlender | [20] | 2018 | Prediction with mean marker expression across references | Brain, pyramidal neurons, stem cells, immune cells, blood cells |
xCell | [39] | 2017 | Linear scaling of marker enrichment scores | Immune, stem, epithelial, and tumor cells |
EPIC | [40] | 2017 | Renormalization of reference markers by cell scale factors, quantification of unknown types | Cancer and blood |
MCP-counter | [41] | 2016 | Cell type amount scoring for heterogeneous tissues, numerous cell types, and multiple clinical conditions | Immune, stromal, and tumor cells and cell lines |
TIMER | [42] | 2016 | Batch effects removal form tumor purity markers; constrained least squares with orthogonal validation | Multiple tumor types |
CIBERSORT | [43] | 2015 | Machine learning-based dimension reduction and permutation optimization | Blood |
DCQ | [44] | 2014 | Whole transcriptome regularized regression followed by ensemble selection, with focus on cell surface marker genes | Lung and immune cells |
DeconRNASeq | [45] | 2013 | Linear modeling, non-negative least squares, and quadratic programming | Brain, heart, skeletal muscle, lung and liver |