Efficient Inference for Sparse Latent Variable Models of Transcriptional Regulation

Zhenwen Dai, Mudassar Iqbal, Neil D. Lawrence, Magnus Rattray
Bioinformatics, 23:3776-3783, 2017.

Abstract

Motivation Regulation of gene expression in prokaryotes involves complex co-regulatory mechanisms involving large numbers of transcriptional regulatory proteins and their target genes. Uncovering these genome-scale interactions constitutes a major bottleneck in systems biology. Sparse latent factor models, assuming activity of transcription factors (TFs) as unobserved, provide a biologically interpretable modelling framework, integrating gene expression and genome-wide binding data, but at the same time pose a hard computational inference problem. Existing probabilistic inference methods for such models rely on subjective filtering and suffer from scalability issues, thus are not well-suited for realistic genome-scale applications. Results We present a fast Bayesian sparse factor model, which takes input gene expression and binding sites data, either from ChIP-seq experiments or motif predictions, and outputs active TF-gene links as well as latent TF activities. Our method employs an efficient variational Bayes scheme for model inference enabling its application to large datasets which was not feasible with existing MCMC-based inference methods for such models. We validate our method on synthetic data against a similar model in the literature, employing MCMC for inference, and obtain comparable results with a small fraction of the computational time. We also apply our method to large-scale data from Mycobacterium tuberculosis involving ChIP-seq data on 113 TFs and matched gene expression data for 3863 putative target genes. We evaluate our predictions using an independent transcriptomics experiment involving over-expression of TFs. Availability and implementation An easy-to-use Jupyter notebook demo of our method with data is available at https://github.com/zhenwendai/SITAR. Supplementary information Supplementary data are available at Bioinformatics online.

Cite this Paper

BibTeX
@Article{Dai-sparse18, title = {Efficient Inference for Sparse Latent Variable Models of Transcriptional Regulation}, author = {Dai, Zhenwen and Iqbal, Mudassar and Lawrence, Neil D. and Rattray, Magnus}, journal = {Bioinformatics}, pages = {3776--3783}, year = {2017}, volume = {23}, doi = {10.1093/bioinformatics/btx508}, pdf = {https://academic.oup.com/bioinformatics/article-pdf/33/23/3776/22031627/btx508.pdf}, url = {http://inverseprobability.com/publications/efficient-inference-for-sparse-latent-variable-models-of-transcriptional-regulation.html}, abstract = {Motivation Regulation of gene expression in prokaryotes involves complex co-regulatory mechanisms involving large numbers of transcriptional regulatory proteins and their target genes. Uncovering these genome-scale interactions constitutes a major bottleneck in systems biology. Sparse latent factor models, assuming activity of transcription factors (TFs) as unobserved, provide a biologically interpretable modelling framework, integrating gene expression and genome-wide binding data, but at the same time pose a hard computational inference problem. Existing probabilistic inference methods for such models rely on subjective filtering and suffer from scalability issues, thus are not well-suited for realistic genome-scale applications. Results We present a fast Bayesian sparse factor model, which takes input gene expression and binding sites data, either from ChIP-seq experiments or motif predictions, and outputs active TF-gene links as well as latent TF activities. Our method employs an efficient variational Bayes scheme for model inference enabling its application to large datasets which was not feasible with existing MCMC-based inference methods for such models. We validate our method on synthetic data against a similar model in the literature, employing MCMC for inference, and obtain comparable results with a small fraction of the computational time. We also apply our method to large-scale data from Mycobacterium tuberculosis involving ChIP-seq data on 113 TFs and matched gene expression data for 3863 putative target genes. We evaluate our predictions using an independent transcriptomics experiment involving over-expression of TFs. Availability and implementation An easy-to-use Jupyter notebook demo of our method with data is available at https://github.com/zhenwendai/SITAR. Supplementary information Supplementary data are available at Bioinformatics online. } }
Endnote
%0 Journal Article %T Efficient Inference for Sparse Latent Variable Models of Transcriptional Regulation %A Zhenwen Dai %A Mudassar Iqbal %A Neil D. Lawrence %A Magnus Rattray %J Bioinformatics %D 2017 %F Dai-sparse18 %P 3776--3783 %R 10.1093/bioinformatics/btx508 %U http://inverseprobability.com/publications/efficient-inference-for-sparse-latent-variable-models-of-transcriptional-regulation.html %V 23 %X Motivation Regulation of gene expression in prokaryotes involves complex co-regulatory mechanisms involving large numbers of transcriptional regulatory proteins and their target genes. Uncovering these genome-scale interactions constitutes a major bottleneck in systems biology. Sparse latent factor models, assuming activity of transcription factors (TFs) as unobserved, provide a biologically interpretable modelling framework, integrating gene expression and genome-wide binding data, but at the same time pose a hard computational inference problem. Existing probabilistic inference methods for such models rely on subjective filtering and suffer from scalability issues, thus are not well-suited for realistic genome-scale applications. Results We present a fast Bayesian sparse factor model, which takes input gene expression and binding sites data, either from ChIP-seq experiments or motif predictions, and outputs active TF-gene links as well as latent TF activities. Our method employs an efficient variational Bayes scheme for model inference enabling its application to large datasets which was not feasible with existing MCMC-based inference methods for such models. We validate our method on synthetic data against a similar model in the literature, employing MCMC for inference, and obtain comparable results with a small fraction of the computational time. We also apply our method to large-scale data from Mycobacterium tuberculosis involving ChIP-seq data on 113 TFs and matched gene expression data for 3863 putative target genes. We evaluate our predictions using an independent transcriptomics experiment involving over-expression of TFs. Availability and implementation An easy-to-use Jupyter notebook demo of our method with data is available at https://github.com/zhenwendai/SITAR. Supplementary information Supplementary data are available at Bioinformatics online.
RIS
TY - JOUR TI - Efficient Inference for Sparse Latent Variable Models of Transcriptional Regulation AU - Zhenwen Dai AU - Mudassar Iqbal AU - Neil D. Lawrence AU - Magnus Rattray DA - 2017/08/26 ID - Dai-sparse18 VL - 23 SP - 3776 EP - 3783 DO - 10.1093/bioinformatics/btx508 L1 - https://academic.oup.com/bioinformatics/article-pdf/33/23/3776/22031627/btx508.pdf UR - http://inverseprobability.com/publications/efficient-inference-for-sparse-latent-variable-models-of-transcriptional-regulation.html AB - Motivation Regulation of gene expression in prokaryotes involves complex co-regulatory mechanisms involving large numbers of transcriptional regulatory proteins and their target genes. Uncovering these genome-scale interactions constitutes a major bottleneck in systems biology. Sparse latent factor models, assuming activity of transcription factors (TFs) as unobserved, provide a biologically interpretable modelling framework, integrating gene expression and genome-wide binding data, but at the same time pose a hard computational inference problem. Existing probabilistic inference methods for such models rely on subjective filtering and suffer from scalability issues, thus are not well-suited for realistic genome-scale applications. Results We present a fast Bayesian sparse factor model, which takes input gene expression and binding sites data, either from ChIP-seq experiments or motif predictions, and outputs active TF-gene links as well as latent TF activities. Our method employs an efficient variational Bayes scheme for model inference enabling its application to large datasets which was not feasible with existing MCMC-based inference methods for such models. We validate our method on synthetic data against a similar model in the literature, employing MCMC for inference, and obtain comparable results with a small fraction of the computational time. We also apply our method to large-scale data from Mycobacterium tuberculosis involving ChIP-seq data on 113 TFs and matched gene expression data for 3863 putative target genes. We evaluate our predictions using an independent transcriptomics experiment involving over-expression of TFs. Availability and implementation An easy-to-use Jupyter notebook demo of our method with data is available at https://github.com/zhenwendai/SITAR. Supplementary information Supplementary data are available at Bioinformatics online. ER -
APA
Dai, Z., Iqbal, M., Lawrence, N.D. & Rattray, M.. (2017). Efficient Inference for Sparse Latent Variable Models of Transcriptional Regulation. Bioinformatics 23:3776-3783 doi:10.1093/bioinformatics/btx508 Available from http://inverseprobability.com/publications/efficient-inference-for-sparse-latent-variable-models-of-transcriptional-regulation.html.

Related Material