A statistical method for deconvolving mixed cancer transcriptomes which addresses the aforementioned issues in array-based expression data. DeMix can be applied to ongoing biomarker-based clinical studies and to the vast expression datasets previously generated from mixed tumor and stromal cell samples.
Provides a statistical approach for deconvolution of mixed tumour profiles given a set of unmatched normal samples. ISOpure runs in two steps: first, it generates a cancer profile estimation by estimating an average cancer profile and a fraction of cancer in each tumour sample. Second, it allows users to make a patient profile estimation for each patient with specific variations.
A package for deconvolution of heterogeneous tissues based on mRNA-Seq data. DeconRNASeq adopts a globally optimized non-negative decomposition algorithm through quadratic programming for estimating the mixing proportions of distinctive tissue types in next-generation sequencing data. It encapsulates the method in a convenient-to-use format. The independent profile-generating module in DeconRNASeq grants freedom to users, who can combine with other R or Bioconductor packages to perform upstream and downstream analysis of NGS data.
Assists users with unsupervised deconvolution of tumor and stromal mixed expression data. UNDO detects cell-specific marker genes (MGs) located on the scatter radii of mixed gene expressions, estimates cellular proportions in each sample and deconvolutes mixed expressions into cell-specific expression profiles. This method also detects differentially expressed genes (DEGs) without deconvolution.
Allows users to detect significant genes for determining cell types and their stages of development. Keygenes is available both as a web platform and an application divided into three scripts: (i) the first one identifies the 500 most variably expressed genes across a next generation sequencing (NGS) dataset; (ii) the second uses an NGS training set to predict an NGS test set; and (iii) the third uses an NGS training set to predict a microarray test set.
Leverages the relationships between tissues and cell-types. URSA is able to identify specific tissue/cell-type signals present in a given gene expression profile. It permits to automatically annotate samples in public gene expression repositories where most samples are currently lacking tissue/cell-type-specific information. The tool can be used to test and identify possible sample contaminations or resolve cancer samples of unknown primary origin.
Conducts differential expression (DE) analysis using high throughput next-generation RNA-seq read count data generated from contaminated tumor samples that are either matched or unmatched with normal samples, which estimates the proportion of pure tumor cells in each contaminated tumor sample, and provides tumor vs. normal log2-fold change, likelihood ratio test statistic and p-value of DE analysis for each gene. It is demonstrated through simulation studies that contamDE could be much more powerful than the existing methods that ignore the contamination.