Efficient Generation of Transcriptomic Profiles by Random Composite Measurements

Abstract

Summary RNA profiles are an informative phenotype of cellular and tissue states, but can be costly to generate at massive scale. Here, we describe how gene expression levels can be efficiently acquired with random composite measurements – in which abundances are combined in a random weighted sum. We show that the similarity between pairs of expression profiles can be approximated with very few composite measurements; that by leveraging sparse, modular representations of gene expression we can use random composite measurements to recover high-dimensional gene expression levels (with 100 times fewer measurements than genes); and that it is possible to blindly recover gene expression from composite measurements, even without access to training data. Our results suggest new compressive modalities as a foundation for massive scaling in high-throughput measurements, and new insights into the interpretation of high-dimensional data.

Improving Deconvolution Methods in Biology through Open Innovation Competitions: An Application to the Connectivity Map

“…In the context of biomedical research, a recurring issue is how to isolate signals of distinct populations (cell types, tissues, and genes) from composite measures obtained by a single analyte or sensor. This problem often stems from the prohibitive cost of profiling each population separately [1,2] and has important implications for the analysis of transcriptional data in mixed samples [3,4,5,6], single-cell data [7], and the study of cell dynamics [8],…”

CoGAPS 3: Bayesian non-negative matrix factorization for single-cell analysis with asynchronous updates and sparse data structures

“…Non-negative matrix factorization (NMF) techniques have emerged as powerful tools to identify the cellular and molecular features that are associated with distinct biological processes from single cell data (Cleary et al, 2017;Zhu et al, 2017;Clark et al, 2019;Welch et al, 2019;Kotliar et al, 2019;Duren et al, 2018). Bayesian factorization approaches can mitigate local optima and leverage prior distributions to encode biological structure in the features Stein-O'Brien et al, 2018).…”

Coexpression enables multi-study cellular trajectories of development and disease

“…This problem has motivated techniques for 14 computational batch effect correction [9][10][11][12][13][14][15] , but existing approaches integrate experiments using 15 transformations that obscure the biological relevance of individual data values, making it 16 difficult for downstream analyses to interpret the transformed result. Existing integrative 17 algorithms also aim to minimize inter-study variation, thus removing relevant differences that 18 would otherwise be useful to biological researchers. 19 To enable robust, consortium-scale scRNA-seq analysis, we reasoned that scRNA-seq 20 analysis of groups of cells in the gene coexpression space (captures the similarity of gene 21 expression changes between pairs of genes), rather than single cells in the gene expression space 22 (focuses on the expression patterns of individual genes), would be a more favorable paradigm.…”

“…measurements; not only are many coexpression measures (for example, Pearson correlation) 25 robust to affine transformation, some evidence suggests that gene coexpression and information 26 redundancy underlie cross-study replicability of single-cell experiments [16][17][18] . Coexpression also 27 provides a rich feature space with directly meaningful values that capture pairwise dependencies 28 among genes, allowing for graph-theoretic analysis of gene coexpression networks.…”