Dataset Information

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

ABSTRACT:

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.

SUBMITTER: Stein-O'Brien GL

PROVIDER: S-EPMC5860188 | biostudies-literature | 2017 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

Stein-O'Brien Genevieve L GL Carey Jacob L JL Lee Wai Shing WS Considine Michael M Favorov Alexander V AV Flam Emily E Guo Theresa T Li Sijia S Marchionni Luigi L Sherman Thomas T Sivy Shawn S Gaykalova Daria A DA McKay Ronald D RD Ochs Michael F MF Colantuoni Carlo C Fertig Elana J EJ

Bioinformatics (Oxford, England) 20170601 12

<h4>Summary</h4>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. ...[more]

PMID: 28174896

Similar Datasets

Project description:BackgroundIn spite of the emergence of RNA sequencing (RNA-seq), microarrays remain in widespread use for gene expression analysis in the clinic. There are over 767,000 RNA microarrays from human samples in public repositories, which are an invaluable resource for biomedical research and personalized medicine. The absolute gene expression analysis allows the transcriptome profiling of all expressed genes under a specific biological condition without the need of a reference sample. However, the background fluorescence represents a challenge to determine the absolute gene expression in microarrays. Given that the Y chromosome is absent in female subjects, we used it as a new approach for absolute gene expression analysis in which the fluorescence of the Y chromosome genes of female subjects was used as the background fluorescence for all the probes in the microarray. This fluorescence was used to establish an absolute gene expression threshold, allowing the differentiation between expressed and non-expressed genes in microarrays.MethodsWe extracted the RNA from 16 children leukocyte samples (nine males and seven females, ages 6-10 years). An Affymetrix Gene Chip Human Gene 1.0 ST Array was carried out for each sample and the fluorescence of 124 genes of the Y chromosome was used to calculate the absolute gene expression threshold. After that, several expressed and non-expressed genes according to our absolute gene expression threshold were compared against the expression obtained using real-time quantitative polymerase chain reaction (RT-qPCR).ResultsFrom the 124 genes of the Y chromosome, three genes (DDX3Y, TXLNG2P and EIF1AY) that displayed significant differences between sexes were used to calculate the absolute gene expression threshold. Using this threshold, we selected 13 expressed and non-expressed genes and confirmed their expression level by RT-qPCR. Then, we selected the top 5% most expressed genes and found that several KEGG pathways were significantly enriched. Interestingly, these pathways were related to the typical functions of leukocytes cells, such as antigen processing and presentation and natural killer cell mediated cytotoxicity. We also applied this method to obtain the absolute gene expression threshold in already published microarray data of liver cells, where the top 5% expressed genes showed an enrichment of typical KEGG pathways for liver cells. Our results suggest that the three selected genes of the Y chromosome can be used to calculate an absolute gene expression threshold, allowing a transcriptome profiling of microarray data without the need of an additional reference experiment.DiscussionOur approach based on the establishment of a threshold for absolute gene expression analysis will allow a new way to analyze thousands of microarrays from public databases. This allows the study of different human diseases without the need of having additional samples for relative expression experiments.

Project description:Chronic motor impairments are a leading cause of disability after stroke. Previous studies have associated motor outcomes with the degree of damage to predefined structures in the motor system, such as the corticospinal tract. However, such theory-based approaches may not take full advantage of the information contained in clinical imaging data. The present study uses data-driven approaches to model chronic motor outcomes after stroke and compares the accuracy of these associations to previously-identified theory-based biomarkers. Using a cross-validation framework, regression models were trained using lesion masks and motor outcomes data from 789 stroke patients from the Enhancing NeuroImaging Genetics through Meta Analysis (ENIGMA) Stroke Recovery Working Group. Using the explained variance metric to measure the strength of the association between chronic motor outcomes and imaging biomarkers, we compared theory-based biomarkers, like lesion load to known motor tracts, to three data-driven biomarkers: lesion load of lesion-behaviour maps, lesion load of structural networks associated with lesion-behaviour maps, and measures of regional structural disconnection. In general, data-driven biomarkers had stronger associations with chronic motor outcomes accuracy than theory-based biomarkers. Data-driven models of regional structural disconnection performed the best of all models tested (R 2 = 0.210, P < 0.001), performing significantly better than the theory-based biomarkers of lesion load of the corticospinal tract (R 2 = 0.132, P < 0.001) and of multiple descending motor tracts (R 2 = 0.180, P < 0.001). They also performed slightly, but significantly, better than other data-driven biomarkers including lesion load of lesion-behaviour maps (R 2 = 0.200, P < 0.001) and lesion load of structural networks associated with lesion-behaviour maps (R 2 = 0.167, P < 0.001). Ensemble models - combining basic demographic variables like age, sex, and time since stroke - improved the strength of associations for theory-based and data-driven biomarkers. Combining both theory-based and data-driven biomarkers with demographic variables improved predictions, and the best ensemble model achieved R 2 = 0.241, P < 0.001. Overall, these results demonstrate that out-of-sample associations between chronic motor outcomes and data-driven imaging features, particularly when lesion data is represented in terms of structural disconnection, are stronger than associations between chronic motor outcomes and theory-based biomarkers. However, combining both theory-based and data-driven models provides the most robust associations.

Dataset Information

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

Summary

Availability and implementation

Contact

Supplementary information

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets