DAISM-DNNXMBD: Highly accurate cell type proportion estimation with in silico data augmentation and deep neural networks

Lin, Yating; Li, Haojun; Xu, Xiao; Zhang, Lei; Wang, Kejia; Zhao, Jingbo; Wang, Minshu; Zheng, Frank; Zhang, Minwei; Yang, Wenxian; Han, Jiahuai; Yu, Rongshan

doi:10.1016/j.patter.2022.100440

Cited by 16 publications

(29 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CIBERSORTx uses ν-SVR to simultaneously solve for all fractional abundances relating admixed and purified expression profiles using a signature matrix of ∼525 differentially expressed genes spanning 22 immune cells types (LM22). 19, 32 Aginome-XMU, published subsequent to the Challenge 33 , utilizes a neural network composed of an input layer, five fully-connected hidden layers, and an output layer (Supplemental Methods; https://github.com/xmuyulab/DCTD_Team_Aginome-XMU; Table S12 ). The network effectively applies feature selection automatically and was trained here using synthetic admixtures.…”

Section: Resultsmentioning

confidence: 99%

“…Finally, as suggested in reported results, 33 it may be challenging to produce a “one-size-fits-all” prediction model that consistently performs well across datasets of different sources due to the existence of batch effects and technical variations from different experiment sites. A better solution seems to be training a “dataset-specific prediction model”.…”

Section: Methodsmentioning

confidence: 92%

“…In addition, we also used a previously-described 33 in silico mixing method to expand a large number of training samples through mixing RNA-seq dataset SDY67 with ground truth and scRNA-seq data.…”

Section: Methodsmentioning

confidence: 99%

“…The detailed description of this deep learning-based model have been published. 33 In addition to the deep learning-based method, in our submission we use an ensemble method to further improve the prediction results. Specifically, we noticed that summarization method (MCP-counter 21 ) performs better on neutrophils, fibroblasts, and endothelial cells on our testing sets.…”

Section: Sub-challengementioning

confidence: 99%

See 3 more Smart Citations

Community assessment of methods to deconvolve cellular composition from bulk gene expression

White

Reyniès

Newman

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Deconvolution methods infer levels of immune and stromal infiltration from bulk expression of tumor samples. These methods allow projection of characteristics of the tumor microenvironment, known to affect patient outcome and therapeutic response, onto the millions of bulk transcriptional profiles in public databases, many focused on uniquely valuable and clinically-annotated cohorts. Despite the wide development of such methods, a standardized dataset with ground truth to evaluate their performance has been lacking. We generated and sequenced in vitro and in silico admixtures of tumor, immune, and stromal cells and used them as ground truth in a community-wide DREAM Challenge that provided an objective, unbiased assessment of six widely-used published deconvolution methods and of 22 new analytical approaches developed by international teams. Our results demonstrate that existing methods predict many cell types well, while team-contributed methods highlight the potential to resolve functional states of T cells that were either not covered by published reference signatures or estimated poorly by some published methods. Our assessment and the open-source implementations of top-performing methods will allow researchers to apply the deconvolution approach most appropriate to querying their cell type of interest. Further, our publicly-available admixed and purified expression profiles will be a valuable resource to those developing deconvolution methods, including in non-malignant settings involving immune cells.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 92%

Section: Methodsmentioning

confidence: 99%

Section: Sub-challengementioning

confidence: 99%

See 2 more Smart Citations

Community assessment of methods to deconvolve cellular composition from bulk gene expression

White

Reyniès

Newman

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We then detail steps to train a dataset-specific deep neural network (DNN) model and cell type proportion estimation using the trained model. For complete details on the use and execution of this protocol, please refer to Lin et al. (2022) .…”

mentioning

confidence: 99%

Protocol to estimate cell type proportions from bulk RNA-seq using DAISM-DNNXMBD

Lin

et al. 2022

STAR Protocols

Self Cite

View full text Add to dashboard Cite

Data-driven bioinformatics to disentangle cells within a tissue microenvironment

Tegnér

2022

Trends in Cell Biology

View full text Add to dashboard Cite

DAISM-DNNXMBD: Highly accurate cell type proportion estimation with in silico data augmentation and deep neural networks

Cited by 16 publications

References 42 publications

Community assessment of methods to deconvolve cellular composition from bulk gene expression

Community assessment of methods to deconvolve cellular composition from bulk gene expression

Protocol to estimate cell type proportions from bulk RNA-seq using DAISM-DNNXMBD

Data-driven bioinformatics to disentangle cells within a tissue microenvironment

Contact Info

Product

Resources

About