Automated deconvolution of structured mixtures from heterogeneous tumor genomic data
“…TCGA is a rich resource for tumor RNA-seq data which has been deconvoluted into cell types by multiple published methods: xCell [26] (64 cell types), which defines gene sets in a pure population and ranks the expression of these in the sample; TIMER [30] (six tumorinfiltrating immune cell types including B cells, CD4 T cells, CD8 T cells, neutrophils, macrophages, and dendritic cells), which uses a signature gene matrix from bulk expression; and others [73][74][75]. We downloaded xCell [26] and TIMER [30] cell fraction data for individuals with KIRC data from TCGA.…”
Section: Deconvoluted Cell Fractions In Tcga Datamentioning
“…TCGA is a rich resource for tumor RNA-seq data which has been deconvoluted into cell types by multiple published methods: xCell [26] (64 cell types), which defines gene sets in a pure population and ranks the expression of these in the sample; TIMER [30] (six tumorinfiltrating immune cell types including B cells, CD4 T cells, CD8 T cells, neutrophils, macrophages, and dendritic cells), which uses a signature gene matrix from bulk expression; and others [73][74][75]. We downloaded xCell [26] and TIMER [30] cell fraction data for individuals with KIRC data from TCGA.…”
Section: Deconvoluted Cell Fractions In Tcga Datamentioning
“…TCGA is a rich resource for tumor RNA-seq data which has been deconvoluted into cell types by multiple published methods: xCell Aran et al (2017) (64 cell types), which defines gene sets in a pure population and ranks the expression of these in the sample; TIMER Li et al (2016) (six tumor-infiltrating immune cell types including B cells, CD4 T cells, CD8 T cells, neutrophils, macrophages, and dendritic cells), which uses a signature gene matrix from bulk expression; and others Wang et al (2018b); Roman et al (2017); Charoentong et al (2017). We downloaded xCell Aran et al (2017) and TIMER Li et al (2016) cell fraction data for individuals with KIRC data from TCGA.…”
Section: Deconvoluted Cell Fractions In Tcga Datamentioning
“…One of the key issues occurs when classical phylogenetic approaches require a priori on the number of sub-clones, which is an unknown parameter for cancer sequencing data. To overcome such issues, the model-free methods often focus on the clonal structures with the maximum likelihood on global variant allelic frequencies [3, 34–38]: THetA designed a convex optimization algorithm to solve the maximum likelihood mixture decomposition, which optimized the multinomial probability [34]. PhyloSub proposed a series of topological constraint rules to limit the possible phylogenies that were able to explain the frequency changes [35].…”
Section: Introductionmentioning
“…TITAN established a graphical model to estimate sub-populations based on copy number alterations and loss of heterozygosity events [37]. Automate learning was also incorporated for deconvolution of genomic mixtures, where the RNA expression data was involved in addition to improve the performance [38]. In general, there is no clear boundary between the two categories, and several comprehensive reviews compared the advantages among the existing approaches [26, 33].…”
Section: Introductionmentioning
See 1 more Smart Citation