Combinatorial prediction of marker panels from single-cell transcriptomic data

Delaney, Conor P.; Schnell, Alexandra; Cammarata, Louis; Yao-Smith, Aaron; Regev, Aviv; Kuchroo, Vijay K.; Singer, Meromit

doi:10.1101/655753

Cited by 15 publications

(32 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We classified circulating CD8 T cells populations based on sharing of TCRs expanded in the tumor. The transcriptomes of individual cells with this annotation were used as the input of COMET (Combinatorial Marker Detection from Single-Cell Transcriptomic data, (Delaney et al, 2019), a tool that uses the minimal hypergeometric test to rank single and paired gene marker panels from a curated list of surface-expressed genes (Chihara et al, 2018) for their ability to identify predefined populations (Fig S3A). Of fourteen candidate gene marker pairs identified in five out of seven patients, the KLRD1-CD74 module had the highest average rank and a highly significant q value in most patients (Fig S3B).…”

Section: Resultsmentioning

confidence: 99%

Circulating Clonally Expanded T Cells Reflect Functions of Tumor Infiltrating T Cells

Lucca

Axisa

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Understanding the relationship between tumor and peripheral immune environments could allow longitudinal immune monitoring in cancer. Here, we examined whether T cells that share the same TCRab and are found in both tumor and blood can be interrogated to gain insight into the ongoing tumor T cell response. Paired transcriptome and TCRab repertoire of circulating and tumor-infiltrating T cells were analyzed from matched tumor and blood from patients with metastatic melanoma at the single cell level. We found that in circulating T cells matching clonally expanded tumor-infiltrating T cells (circulating TILs), gene signatures of effector functions, but not terminal exhaustion, reflect those observed in the tumor. In contrast, features of exhaustion are displayed predominantly by T cells present only in tumor. Finally, genes associated with a high degree of blood-tumor TCR sharing were overexpressed in tumor tissue after immunotherapy. These data demonstrate that circulating TILs, identified by TCRs shared with T cells in tumors, have unique transcriptional expression patterns that may have utility for the interrogation of T cell function in cancer immunotherapy.

show abstract

Section: Resultsmentioning

confidence: 99%

Circulating Clonally Expanded T Cells Reflect Functions of Tumor Infiltrating T Cells

Lucca

Axisa

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…a SCA cluster retaining the cell organization of at least one of the reference clusters, should have both QCF and QCM ³ 0.5. A frequency matrix for the latent space representations is also built and it is used as input for COMET [20], which is a software able to identify marker signatures specific of each cluster. Only clusters characterized by QCM and QCF means ³ 0.5 should be used for marker signature detection.…”

Section: Resultsmentioning

confidence: 99%

“…The matrix describing the frequency of the latent space variables is used to extract cluster specific signatures using COMET tool [20], which is also implemented in rCASC. COMET is a computational framework for the identification of candidate marker panels consisting of one or more genes for cell populations of interest identified with single cell RNAseq data.…”

Section: Sca Analysis On a Pbmc Derived Dataset (Seta)mentioning

confidence: 99%

See 1 more Smart Citation

Sparsely-Connected Autoencoder (SCA) for single cell RNAseq data mining

Alessandrì

Cordero

Beccuti

et al. 2020

Preprint

View full text Add to dashboard Cite

Single-cell RNA sequencing (scRNAseq) is an essential tool to investigate cellular heterogeneity. Although scRNAseq has some technical challenges, it would be of great interest being able to disclose biological information out of cell subpopulations, which can be defined by cluster analysis of scRNAseq data. In this manuscript, we evaluated the efficacy of sparsely-connected autoencoder (SCA) as tool for the functional mining of single cells clusters. We show that SCA can be uses as tool to uncover hidden features associated to scRNAseq data. Our approach is strengthened by two metrics, QCF and QCM, which respectively allow to evaluate the ability of SCA to reconstruct a cells cluster and to evaluate the overall quality of the neural network model. Our data indicate that SCA encoded spaces, derived by different experimentally validated data (TFs targets, miRNAs targets, Kinases targets, and cancer-related immune signatures), can be used to grasp single cell clusterspecific functional features. In our implementation, SCA efficacy comes from its ability to reconstruct only specific clusters, thus indicating only those clusters where the SCA encoding space is a key element for cells aggregation. SCA analysis is implemented as module in rCASC framework and it is supported by a GUI to simplify it usage for biologists and medical personnel.

show abstract

“…In two other examples, we used human cells, endocrine Veres et al (2019) and from brain organoids Gray Camp et al (2015), showing that the method works robustly in varied conditions. Other methods to identify important genes from scRNAseq data exist, but most of them aim to find marker genes that can best distinguish different cell types Delaney et al (2019). Conversely, the method we presented is unsupervised and does not rely on cell type annotation.…”

Section: Discussionmentioning

confidence: 99%

Automatic identification of relevant genes from low-dimensional embeddings of single cell RNAseq data

Angerer

Fischer

Theis

et al. 2020

Preprint

View full text Add to dashboard Cite

Dimensionality reduction is a key step in the analysis of single-cell RNA sequencing data and produces a low-dimensional embedding for visualization and as a calculation base for downstream analysis. Nonlinear techniques are most suitable to handle the intrinsic complexity of large, heterogeneous single cell data. With no linear relation between genes and embedding however, there is no way to extract the identity of genes most relevant for any cell's position in the low-dimensional embedding, and thus the underlying process. In this paper, we introduce the concepts of global and local gene relevance to compute an equivalent of principal component analysis loadings for non-linear low-dimensional embeddings. While global gene relevance identifies drivers of the overall embedding, local gene relevance singles out genes that change in small, possibly rare subsets of cells. We apply our method to single-cell RNAseq datasets from different experimental protocols and to different low dimensional embedding techniques, shows our method's versatility to identify key genes for a variety of biological processes. To ensure reproducibility and ease of use, our method is released as part of destiny 3.0, a popular R package for building diffusion maps from single-cell transcriptomic data. It is readily available through Bioconductor.

show abstract

Combinatorial prediction of marker panels from single-cell transcriptomic data

Cited by 15 publications

References 42 publications

Circulating Clonally Expanded T Cells Reflect Functions of Tumor Infiltrating T Cells

Circulating Clonally Expanded T Cells Reflect Functions of Tumor Infiltrating T Cells

Sparsely-Connected Autoencoder (SCA) for single cell RNAseq data mining

Automatic identification of relevant genes from low-dimensional embeddings of single cell RNAseq data

Contact Info

Product

Resources

About