Shallow Sparsely-Connected Autoencoders for Gene Set Projection

Gold, Maxwell P.; LeNail, Alexander; Fraenkel, Ernest

doi:10.1142/9789813279827_0034

Cited by 14 publications

(25 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sparsely-Connected Autoencoder (SCA) SCA encoding/decoding functions consist of a single sparse layer ( Fig. 1, latent space), with connections based on known biological relationships [8,10]. Each node represents a known biological relationship (transcription factors (TFs) targets, miRNA targets (miRNAs), cancer-related immune-signatures (ISs), kinase specific protein targets (Ks), etc.)…”

Section: Resultsmentioning

confidence: 99%

“…and only receives inputs from gene nodes associated with the biological relationship. With respect to the Gold paper [8], which uses gene sets [11], in our implementation the latent space is based on experimentally validated data, TRRUST [12], miRTarBase [13], RegPhos [14] , and a manually curated cancer-based immunesignature (See material and methods). SCA analysis is executed multiple times on a cell dataset, previously partitioned in clusters, using any of the clustering tools implemented in rCASC: tSne+k-mean [15], SIMLR [16], griph [17], scanpy [18] and SHARP [19].…”

Section: Resultsmentioning

confidence: 99%

“…Autoencoder-based approaches have been used to cluster single cell data [4], to impute single cell data [5], data denoising [6], batch correction [7]. Recently, Gold and co-workers [8] have evaluated the use of autoencoders for data interpretation, implementing sparsely-connected autoencoder (SCA) to gene set analysis. SCA uses a single-layer autoencoder with sparse connections (representing known biological relationships) in order to attain a value for each gene set.…”

Section: Introductionmentioning

confidence: 99%

“…SCA uses a single-layer autoencoder with sparse connections (representing known biological relationships) in order to attain a value for each gene set. SCA provides great flexibility for modelling biological phenomena [8]. We made available to the single-cell community a framework, rCASC [9], providing an integrated analysis environment for single-cell RNAseq.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

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

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

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

Alessandrì

Cordero

Beccuti

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, an approach with deep generative modeling for scRNA-Seq data normalization and domain adaptation was recently proposed in [6]. Another approach to gene expression data modelling with autoencoders was presented in [11] -the authors induced the sparsity of network weights by connecting only the genes from the same functional group to the same hidden neuron. This is a step towards interpretability of autoencoder models.…”

Section: Introductionmentioning

confidence: 99%

Style transfer with variational autoencoders is a promising approach to RNA-Seq data harmonization and analysis

Russkikh¹,

Антонец

Shtokalo

et al. 2019

Preprint

View full text Add to dashboard Cite

The transcriptomic data is being frequently used in the research of biomarker genes of different diseases and biological states. Generally, researchers have data from hundreds, rarely thousands of specimens at hand. In most cases, the proposed candidate biomarker genes and corresponding decision rules fail in prospective research studies, especially for diseases with complex polygenic background. The naive addition of training data usually also does not improve performance due to batch effects, resulting from various discrepancies between different datasets. To get a better understanding of factors underlying the observed gene expression data variation, we applied a style transfer technique. The most of style transfer studies are focused on image data, and, to our knowledge, this is the first attempt to adapt this procedure to gene expression domain. As a style component, there might be used either technical factors of data variance, such as sequencing platform, RNA extraction protocols, or any biological details about the samples which we would like to control (gender, biological state, treatment etc.). The proposed solution is based on Variational Autoencoder artificial neural network. To disentangle the style components, we trained the encoder with discriminator in an adversarial manner. This approach can be useful for both data harmonization and data augmentation -for obtaining semisynthetic samples when the real data is scarce. We demonstrated the applicability of our framework using single cell RNA-Seq data from Mouse Cell Atlas, where we were able to transfer the mammary gland biological state (virgin, pregnancy and involution state)

show abstract

Application of Deep Learning on Single-Cell RNA Sequencing Data Analysis: A Review

Brendel

Bai

et al. 2022

Genomics, Proteomics &Amp; Bioinformatics

View full text Add to dashboard Cite

Single-cell RNA sequencing (scRNA-seq) has become a routinely used technique to quantify the gene expression profile of thousands of single cells simultaneously. Analysis of scRNA-seq data plays an important role in the study of cell states and phenotypes, and has helped elucidate biological processes, such as those occurring during the development of complex organisms, and improved our understanding of disease states, such as cancer, diabetes, and coronavirus disease 2019 (COVID-19). Deep learning, a recent advance of artificial intelligence that has been used to address many problems involving large datasets, has also emerged as a promising tool for scRNA-seq data analysis, as it has a capacity to extract informative and compact features from noisy, heterogeneous, and high-dimensional scRNA-seq data to improve downstream analysis. The present review aims at surveying recently developed deep learning techniques in scRNA-seq data analysis, identifying key steps within the scRNA-seq data analysis pipeline that have been advanced by deep learning, and explaining the benefits of deep learning over more conventional analytic tools. Finally, we summarize the challenges in current deep learning approaches faced within scRNA-seq data and discuss potential directions for improvements in deep learning algorithms for scRNA-seq data analysis.

show abstract

Shallow Sparsely-Connected Autoencoders for Gene Set Projection

Cited by 14 publications

References 29 publications

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

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

Style transfer with variational autoencoders is a promising approach to RNA-Seq data harmonization and analysis

Application of Deep Learning on Single-Cell RNA Sequencing Data Analysis: A Review

Contact Info

Product

Resources

About