Discovering subgroups of patients from DNA copy number data using NMF on compacted matrices

A prognostic model related to necrotizing apoptosis of breast cancer based on biorthogonal constrained depth semi-supervised nonnegative matrix decomposition and single-cell sequencing analysis

Breast cancer (BC) is one of the most common malignant tumours in women, and its prognosis is poor. The prognosis of BC patients can be improved by immunotherapy. However, due to the heterogeneity of BC, the identification of new biomarkers is urgently needed to improve the prognosis of BC patients. Necrotic apoptosis has been shown to play an essential role in many cancers. First, this study proposed a novel clustering algorithm called biorthogonal constrained depth semisupervised nonnegative matrix factorization (DO-DSNMF). The DO-DSNMF algorithm added multilayer nonlinear transformation to the coefficient matrix obtained after decomposition, which was used to mine the nonlinear relationship between samples. In addition, we also added orthogonal constraints on the basis matrix and coefficient matrix to reduce the influence of redundant features and samples on the results. We applied the DO-DSNMF algorithm and analysed the differences in survival and immunity between the subtypes. Then, we used prognosis analysis to construct the prognosis model. Finally, we analysed single cells using single-cell sequencing (scRNA-seq) data from the GSE75688 dataset in the GEO database. We identified two BC subtypes based on the BC transcriptome data in the TCGA database. Immune infiltration analysis showed that the necrotizing apoptosis-related genes of BC were related to various immune cells and immune functions. Necrotizing apoptosis was found to play a role in BC progression and immunity. The role of prognosis-related NRGs in BC was also verified by cell experiments. This study proposed a novel clustering algorithm to analyse BC subtypes and constructed an NRG prognostic model for BC. The prognosis and immune landscape of BC patients were evaluated by this model. The cell experiment supported its role in BC, which provides a potential therapeutic target for the treatment of BC.

Unsupervised Algorithms for Microarray Sample Stratification

The amount of data made available by microarrays gives researchers the opportunity to delve into the complexity of biological systems. However, the noisy and extremely high-dimensional nature of this kind of data poses significant challenges. Microarrays allow for the parallel measurement of thousands of molecular objects spanning different layers of interactions. In order to be able to discover hidden patterns, the most disparate analytical techniques have been proposed. Here, we describe the basic methodologies to approach the analysis of microarray datasets that focus on the task of (sub)group discovery.

Enter the Matrix: Factorization Uncovers Knowledge from Omics

Omics data contain signals from the molecular, physical, and kinetic inter- and intracellular interactions that control biological systems. Matrix factorization (MF) techniques can reveal low-dimensional structure from high-dimensional data that reflect these interactions. These techniques can uncover new biological knowledge from diverse high-throughput omics data in applications ranging from pathway discovery to timecourse analysis. We review exemplary applications of MF for systems-level analyses. We discuss appropriate applications of these methods, their limitations, and focus on the analysis of results to facilitate optimal biological interpretation. The inference of biologically relevant features with MF enables discovery from high-throughput data beyond the limits of current biological knowledge - answering questions from high-dimensional data that we have not yet thought to ask.

PatternMarkers & GWCoGAPS for novel data-driven biomarkers via whole transcriptome NMF

Summary

Non-negative Matrix Factorization (NMF) algorithms associate gene expression with biological processes (e.g. time-course dynamics or disease subtypes). Compared with univariate associations, the relative weights of NMF solutions can obscure biomarkers. Therefore, we developed a novel patternMarkers statistic to extract genes for biological validation and enhanced visualization of NMF results. Finding novel and unbiased gene markers with patternMarkers requires whole-genome data. Therefore, we also developed Genome-Wide CoGAPS Analysis in Parallel Sets (GWCoGAPS), the first robust whole genome Bayesian NMF using the sparse, MCMC algorithm, CoGAPS. Additionally, a manual version of the GWCoGAPS algorithm contains analytic and visualization tools including patternMatcher, a Shiny web application. The decomposition in the manual pipeline can be replaced with any NMF algorithm, for further generalization of the software. Using these tools, we find granular brain-region and cell-type specific signatures with corresponding biomarkers in GTEx data, illustrating GWCoGAPS and patternMarkers ascertainment of data-driven biomarkers from whole-genome data.

Availability and Implementation

PatternMarkers & GWCoGAPS are in the CoGAPS Bioconductor package (3.5) under the GPL license.

Contact

gsteinobrien@jhmi.edu or ccolantu@jhmi.edu or ejfertig@jhmi.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

Identification of hub genes and regulatory factors of glioblastoma multiforme subgroups by RNA-seq data analysis

Glioblastoma multiforme (GBM) is the most common malignant brain tumor. This study aimed to identify the hub genes and regulatory factors of GBM subgroups by RNA sequencing (RNA-seq) data analysis, in order to explore the possible mechanisms responsbile for the progression of GBM. The dataset RNASeqV2 was downloaded by TCGA-Assembler, containing 169 GBM and 5 normal samples. Gene expression was calculated by the reads per kilobase per million reads measurement, and nor malized with tag count comparison. Following subgroup classification by the non-negative matrix factorization, the differentially expressed genes (DEGs) were screened in 4 GBM subgroups using the method of significance analysis of microarrays. Functional enrichment analysis was performed by DAVID, and the protein-protein interaction (PPI) network was constructed based on the HPRD database. The subgroup-related microRNAs (miRNAs or miRs), transcription factors (TFs) and small molecule drugs were predicted with predefined criteria. A cohort of 19,515 DEGs between the GBM and control samples was screened, which were predominantly enriched in cell cycle- and immunoreaction-related pathways. In the PPI network, lymphocyte cytosolic protein 2 (LCP2), breast cancer 1 (BRCA1), specificity protein 1 (Sp1) and chromodomain-helicase-DNA-binding protein 3 (CHD3) were the hub nodes in subgroups 1–4, respectively. Paired box 5 (PAX5), adipocyte protein 2 (aP2), E2F transcription factor 1 (E2F1) and cAMP-response element-binding protein-1 (CREB1) were the specific TFs in subgroups 1–4, respectively. miR-147b, miR-770-5p, miR-220a and miR-1247 were the particular miRNAs in subgroups 1–4, respectively. Natalizumab was the predicted small molecule drug in subgroup 2. In conclusion, the molecular regulatory mechanisms of GBM pathogenesis were distinct in the different subgroups. Several crucial genes, TFs, miRNAs and small molecules in the different GBM subgroups were identified, which may be used as potential markers. However, further experimental validations may be required.