QHQiwen Hu

Qiwen Hu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Qiwen Hu has a total of 13 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Parameter tuning is a key part of dimensionality reduction via deep variational autoencoders for single cell RNA transcriptomics

    by Qiwen Hu, Casey S. Greene

    Single-cell RNA sequencing (scRNA-seq) is a powerful tool to profile the transcriptomes of a large number of individual cells at a high resolution. These data usually contain measurements of gene expression for many genes in thousands or tens of thousands of cells, though some datasets now reach the million-cell mark. Projecting high-dimensional scRNA-seq data into a low dimensional space aids downstream analysis and data visualization. Many recent preprints accomplish this using variational autoencoders (VAE), generative models that learn underlying structure of data by compress it into a constrained, low dimensional space. The low dimensional spaces generated by VAEs have revealed complex patterns and novel biological signals from large-scale gene expression data and drug response predictions. Here, we evaluate a simple VAE approach for gene expression data, Tybalt, by training and measuring its performance on sets of simulated scRNA-seq data. We find a number of counter-intuitive performance features: i.e., deeper neural networks can struggle when datasets contain more observations under some parameter configurations. We show that these methods are highly sensitive to parameter tuning: when tuned, the performance of the Tybalt model, which was not optimized for scRNA-seq data, outperforms other popular dimension reduction approaches – PCA, ZIFA, UMAP and t-SNE. On the other hand, without tuning performance can also be remarkably poor on the same data. Our results should discourage authors and reviewers from relying on self-reported performance comparisons to evaluate the relative value of contributions in this area at this time. Instead, we recommend that attempts to compare or benchmark autoencoder methods for scRNA-seq data be performed by disinterested third parties or by methods developers only on unseen benchmark data that are provided to all participants simultaneously because the potential for performance differences due to unequal parameter tuning is so high.


  • Open Access

    Decomposing cell identity for transfer learning across cellular measurements, platforms, tissues, and species

    by Genevieve L. Stein-O’brien, Brian S. Clark, Thomas Sherman, Christina Zibetti, Qiwen Hu, Rachel Sealfon, Sheng Liu, Jiang Qian, Carlo Colantuoni, Seth Blackshaw, et al.

    New approaches are urgently needed to glean biological insights from the vast amounts of single cell RNA sequencing (scRNA-Seq) data now being generated. To this end, we propose that cell identity should map to a reduced set of factors which will describe both exclusive and shared biology of individual cells, and that the dimensions which contain these factors reflect biologically meaningful relationships across different platforms, tissues and species. To find a robust set of dependent factors in large-scale scRNA- Seq data, we developed a Bayesian non-negative matrix factorization (NMF) algorithm, scCoGAPS. Application of scCoGAPS to scRNA-Seq data obtained over the course of mouse retinal development identified gene expression signatures for factors associated with specific cell types and continuous biological processes. To test whether these signatures are shared across diverse cellular contexts, we developed projectR to map biologically disparate datasets into the factors learned by scCoGAPS. Because projecting these dimensions preserve relative distances between samples, biologically meaningful relationships/factors will stratify new data consistent with their underlying processes, allowing labels or information from one dataset to be used for annotation of the other—a machine learning concept called transfer learning. Using projectR, data from multiple datasets was used to annotate latent spaces and reveal novel parallels between developmental programs in other tissues, species and cellular assays. Using this approach we are able to transfer cell type and state designations across datasets to rapidly annotate cellular features in a new dataset without a priori knowledge of their type, identify a species-specific signature of microglial cells, and identify a previously undescribed subpopulation of neurosecretory cells within the lung. Together, these algorithms define biologically meaningful dimensions of cellular identity, state, and trajectories that persist across technologies, molecular features, and species. GRAPHICAL


  • Open Access

    Parameter tuning is a key part of dimensionality reduction via deep variational autoencoders for single cell RNA transcriptomics

    by Qiwen Hu, Casey S. Greene

    Single-cell RNA sequencing (scRNA-seq) is a powerful tool to profile the transcriptomes of a large number of individual cells at a high resolution. These data usually contain measurements of gene expression for many genes in thousands or tens of thousands of cells, though some datasets now reach the million-cell mark. Projecting high-dimensional scRNA-seq data into a low dimensional space aids downstream analysis and data visualization. Many recent preprints accomplish this using variational autoencoders (VAE), generative models that learn underlying structure of data by compress it into a constrained, low dimensional space. The low dimensional spaces generated by VAEs have revealed complex patterns and novel biological signals from large-scale gene expression data and drug response predictions. Here, we evaluate a simple VAE approach for gene expression data, Tybalt, by training and measuring its performance on sets of simulated scRNA-seq data. We find a number of counter-intuitive performance features: i.e., deeper neural networks can struggle when datasets contain more observations under some parameter configurations. We show that these methods are highly sensitive to parameter tuning: when tuned, the performance of the Tybalt model, which was not optimized for scRNA-seq data, outperforms other popular dimension reduction approaches – PCA, ZIFA, UMAP and t-SNE. On the other hand, without tuning performance can also be remarkably poor on the same data. Our results should discourage authors and reviewers from relying on self-reported performance comparisons to evaluate the relative value of contributions in this area at this time. Instead, we recommend that attempts to compare or benchmark autoencoder methods for scRNA-seq data be performed by disinterested third parties or by methods developers only on unseen benchmark data that are provided to all participants simultaneously because the potential for performance differences due to unequal parameter tuning is so high.


  • Open Access

    Specific histone modifications associate with alternative exon selection during mammalian development

    by Qiwen Hu, Casey Greene, Elizabeth Heller

    Alternative splicing (AS) is frequent during early mouse embryonic development. Specific histone post-translational modifications (hPTMs) have been shown to regulate exon splicing by either directly recruiting splice machinery or indirectly modulating transcriptional elongation. In this study, we hypothesized that hPTMs regulate expression of alternatively spliced genes for specific processes during differentiation. To address this notion, we applied an innovative machine learning approach to relate global hPTM enrichment to AS regulation during mammalian tissue development. We found that specific histone modifications, H3K36me3 and H3K4me1, play a dominant role in skipped exon selection among all the tissues and developmental time points examined. In addition, we used iterative random forest model to identify interactions of several hPTMs that associated with skipped exon selection during tissue development. Collectively, our data demonstrated a link between hPTMs and alternative splicing which will drive further experimental studies on the functional relevance of these modifications to alternative splicing.