MINT: A multivariate integrative method to identify reproducible molecular signatures across independent experiments and platforms

Abstract

Background: Molecular signatures identified from high-throughput transcriptomic studies often have poor reliability and fail to reproduce across studies. One solution is to combine independent studies into a single integrative analysis, additionally increasing sample size. However, the different protocols and technological platforms across transcriptomic studies produce unwanted systematic variation that strongly confounds the integrative analysis results. When studies aim to discriminate an outcome of interest, the common approach is a sequential two-step procedure; unwanted systematic variation removal techniques are applied prior to classification methods. Results: To limit the risk of overfitting and over-optimistic results of a two-step procedure, we developed a novel multivariate integration method, MINT, that simultaneously accounts for unwanted systematic variation and identifies predictive gene signatures with greater reproducibility and accuracy. In two biological examples on the classification of three human cell types and four subtypes of breast cancer, we combined high-dimensional microarray and RNA-seq data sets and MINT identified highly reproducible and relevant gene signatures predictive of a given phenotype. MINT led to superior classification and prediction accuracy compared to the existing sequential two-step procedures. Conclusions: MINT is a powerful approach and the first of its kind to solve the integrative classification framework in a single step by combining multiple independent studies. MINT is computationally fast as part of the mixOmics R CRAN package, available at

scRNA-seq mixology: towards better benchmarking of single cell RNA-seq analysis methods

“…None of the methods proposed so far have been compared on well-designed benchmark datasets generated using multiple protocols. We used the single cell (sc CELseq2, sc 10X, sc Dropseq and sc 10x 5cl) and RNA mixture (RNAmix CELseq2 and RNAmix Sortseq) experiments to compare state-of-art methods including MNNs [13], Scanorama [16], scMerge [27], Seurat [4] and MINT [37]. The 5 cell line CEL-seq2 datasets (sc CELseq2 5cl p1, sc CELseq2 5cl p2 and sc CELseq2 5cl p3) were excluded due to their high doublet rates (Supplementary Figure 3B).…”

“…MNNs, Scanorama and scMerge generate batch-corrected data which can then be analyzed using other downstream analysis tools, while Diagonal Canonical Correlation Analysis combined with Dynamic Time Warping from Seurat and MINT [37] output a low-dimensional representation of the data. MINT includes an embedded gene selection procedure whilst projecting the data into a lower dimensional space.…”

The chemical profiling of fatty acids during the brewing process

“…Statistical analysis, data treatment and preparation of graphs was performed using R, version 3.4.0 (R Foundation for Statistical Computing, Vienna, Austria) with the muma , mixOmics and pheatmap packages . Computed concentrations of FA from two independent batches (from years 2013 and 2016) were integrated by a multivariate integrative method (MINT) to minimize systematic unwanted variation between batches. Data were then scaled by z ‐score and visualized in a heat map where the columns were clustered by hierarchical clustering based on Euclidean distances and the Ward method.…”

Mitochondrial translation and dynamics synergistically extend lifespan inC. elegansthrough HLH-30

“…Protein data were generated from the peptide matrix using MSstats v3.12.0. Differential expression was determined using either a Student's t-test (with p values corrected for multiple testing using the Benjamini-Hochberg false discovery rate, presented for reference) or partial least squares discriminant analysis (PLS-DA) with mixomics (Rohart et al, 2017) setting a variable of importance (VIP) score of greater than 1 as significant. The DIA data and the assembled library file are available on PRIDE under the ID PXD009256.…”