Dataset Information

Probabilistic cell-type assignment of single-cell RNA-seq for tumor microenvironment profiling.

ABSTRACT: Single-cell RNA sequencing has enabled the decomposition of complex tissues into functionally distinct cell types. Often, investigators wish to assign cells to cell types through unsupervised clustering followed by manual annotation or via 'mapping' to existing data. However, manual interpretation scales poorly to large datasets, mapping approaches require purified or pre-annotated data and both are prone to batch effects. To overcome these issues, we present CellAssign, a probabilistic model that leverages prior knowledge of cell-type marker genes to annotate single-cell RNA sequencing data into predefined or de novo cell types. CellAssign automates the process of assigning cells in a highly scalable manner across large datasets while controlling for batch and sample effects. We demonstrate the advantages of CellAssign through extensive simulations and analysis of tumor microenvironment composition in high-grade serous ovarian cancer and follicular lymphoma.

SUBMITTER: Zhang AW

PROVIDER: S-EPMC7485597 | biostudies-literature | 2019 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Probabilistic cell-type assignment of single-cell RNA-seq for tumor microenvironment profiling.

Zhang Allen W AW O'Flanagan Ciara C Chavez Elizabeth A EA Lim Jamie L P JLP Ceglia Nicholas N McPherson Andrew A Wiens Matt M Walters Pascale P Chan Tim T Hewitson Brittany B Lai Daniel D Mottok Anja A Sarkozy Clementine C Chong Lauren L Aoki Tomohiro T Wang Xuehai X Weng Andrew P AP McAlpine Jessica N JN Aparicio Samuel S Steidl Christian C Campbell Kieran R KR Shah Sohrab P SP

Nature methods 20190909 10

Single-cell RNA sequencing has enabled the decomposition of complex tissues into functionally distinct cell types. Often, investigators wish to assign cells to cell types through unsupervised clustering followed by manual annotation or via 'mapping' to existing data. However, manual interpretation scales poorly to large datasets, mapping approaches require purified or pre-annotated data and both are prone to batch effects. To overcome these issues, we present CellAssign, a probabilistic model th ...[more]

PMID: 31501550

Similar Datasets

Project description:Background: The challenge of systemic treatment for hepatocellular carcinoma (HCC) stems from the development of drug resistance, primarily driven by the interplay between cancer stem cells (CSCs) and the tumor microenvironment (TME). However, there is a notable dearth of comprehensive research investigating the crosstalk between CSCs and stromal cells or immune cells within the TME of HCC. Methods: We procured single-cell RNA sequencing (scRNA-Seq) data from 16 patients diagnosed with HCC. Employing meticulous data quality control and cell annotation procedures, we delineated distinct CSCs subtypes and performed multi-omics analyses encompassing metabolic activity, cell communication, and cell trajectory. These analyses shed light on the potential molecular mechanisms governing the interaction between CSCs and the TME, while also identifying CSCs' developmental genes. By combining these developmental genes, we employed machine learning algorithms and RT-qPCR to construct and validate a prognostic risk model for HCC. Results: We successfully identified CSCs subtypes residing within malignant cells. Through meticulous enrichment analysis and assessment of metabolic activity, we discovered anomalous metabolic patterns within the CSCs microenvironment, including hypoxia and glucose deprivation. Moreover, CSCs exhibited aberrant activity in signaling pathways associated with lipid metabolism. Furthermore, our investigations into cell communication unveiled that CSCs possess the capacity to modulate stromal cells and immune cells through the secretion of MIF or MDK, consequently exerting regulatory control over the TME. Finally, through cell trajectory analysis, we found developmental genes of CSCs. Leveraging these genes, we successfully developed and validated a prognostic risk model (APCS, ADH4, FTH1, and HSPB1) with machine learning and RT-qPCR. Conclusions: By means of single-cell multi-omics analysis, this study offers valuable insights into the potential molecular mechanisms governing the interaction between CSCs and the TME, elucidating the pivotal role CSCs play within the TME. Additionally, we have successfully established a comprehensive clinical prognostic model through bulk RNA-Seq data.

Dataset Information

Probabilistic cell-type assignment of single-cell RNA-seq for tumor microenvironment profiling.

Publications

Probabilistic cell-type assignment of single-cell RNA-seq for tumor microenvironment profiling.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets