Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

Abstract

Single-cell RNA-sequencing (RNA-seq) has become a powerful tool to study biologically significant characteristics at explicitly high resolution. However, its application on emerging data is currently limited by its intrinsic techniques. Though many methods have been proposed to analyze bulk data using single-cell profile as a reference, they are limited on the interpretability, processing speed, and data size requirement. Here, we introduce Tissue-AdaPtive autoEncoder (TAPE), a deep learning method connecting bulk RNA-seq and single-cell RNA-seq, to achieve precise prediction in short time. By constructing an interpretable decoder and training under a unique scheme, TAPE can predict cell-type fractions and cell-type-specific gene expression tissue-adaptively. Compared with existing methods on several benchmarking datasets, TAPE is more accurate (up to 40% improvement on the real bulk data) and faster. Thus, it is sensitive enough to provide biologically meaningful predictions. For example, only TAPE can predict the tendency of increasing monocytes-to-lymphocytes ratio in COVID-19 patients from mild to serious symptoms, of which estimated indices are consistent with laboratory data. More importantly, through the analysis of clinical data, TAPE shows its ability to predict cell-type-specific gene expression profiles with biological significance. Combining with single-sample gene set enrichment analysis (ssGSEA), TAPE also provides valuable clues to investigate the immune response in different virus-infected patients. We believe that TAPE will enable and accelerate the precise analysis of high-throughput clinical data in a wide range.

Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis

“…The open source implementation of TAPE is available at https://github. com/poseidonchan/TAPE 59 , and the experiments conducted to produce the main results of this article are also stored in this repository. The documentation of TAPE is published at https://sctape.…”