Scalable preprocessing for sparse scRNA-seq data exploiting prior knowledge

Abstract

Single cell RNA-seq (scRNA-seq) experiments can provide a wealth of information about heterogeneous, multi-cellular systems. However, this information has to be inferred computationally from sequencing reads which constitute a sparse and noisy sub-sampling of the actual cellular transcriptomes. Here we present UNCURL (https://github.com/mukhes3/UNCURL_release), a unified framework for scRNA-seq data visualization, cell type identification and lineage estimation that explicitly accounts for the sequencing process. The main algorithmic novelty is a non-negative matrix factorization method that uses knowledge of the distribution resulting from the sequencing process to more accurately model the underlying cell state matrix. We also develop a systematic way for incorporating prior biological information such as bulk RNA expression profiles into the cell state matrix. We find that UNCURL dramatically improves performance over state-of-the-art methods both in the absence and presence of prior knowledge. Finally we demonstrate that using UNCURL as a data preprocessing tool significantly improves the performance of existing scRNA-seq analysis algorithms.

Learning interpretable cellular and gene signature embeddings from single-cell transcriptomic data

“…They are also non-transferable, meaning that the knowledge learned from one dataset cannot be easily transferred through model parameter sharing to benefit the modeling of another dataset. NMF approaches such as UNCURL [30] works only with one scRNA-seq dataset. LIGER [9] uses integrative NMF to jointly factorize multiple scRNA-seq matrices across conditions using genes as the common axis, linking cells from different conditions by a common set of latent factors also known as metagenes.…”

Big data analytics in single‐cell transcriptomics: Five grand opportunities

“…Several “imputation” methods have been developed to tackle the dropouts (Mukherjee, Zhang, Fan, Seelig, & Kannan, 2018; Peng, Zhu, Yin, & Tan, 2019). To impute dropouts, matrix completion based imputation for single‐cell RNA‐seq data (McImpute) uses a technique based on low‐rank matrix completion of sparse single‐cell expression data (Mongia, Sengupta, & Majumdar, 2019).…”