Background: Unsupervised compression algorithms applied to gene expression data extract latent or hidden signals representing technical and biological sources of variation. However, these algorithms require a user to select a biologically appropriate latent space dimensionality. In practice, most researchers fit a single algorithm and latent dimensionality. We sought to determine the extent by which selecting only one fit limits the biological features captured in the latent representations and, consequently, limits what can be discovered with subsequent analyses. Results: We compress gene expression data from three large datasets consisting of adult normal tissue, adult cancer tissue, and pediatric cancer tissue. We train many different models across a large range of latent space dimensionalities and observe various performance differences. We identify more curated pathway gene sets significantly associated with individual dimensions in denoising autoencoder and variational autoencoder models trained using an intermediate number of latent dimensionalities. Combining compressed features across algorithms and dimensionalities captures the most pathway-associated representations. When trained with different latent dimensionalities, models learn strongly associated and generalizable biological representations including sex, neuroblastoma MYCN amplification, and cell types. Stronger signals, such as tumor type, are best captured in models trained at lower dimensionalities, while more subtle signals such as pathway activity are best identified in models trained with more latent dimensionalities. Conclusions: There is no single best latent dimensionality or compression algorithm for analyzing gene expression data. Instead, using features derived from different compression models across multiple latent space dimensionalities enhances biological representations.

Open, collaborative research is a powerful paradigm that can immensely strengthen the scientific process by integrating broad and diverse expertise. However, traditional research and multi-author writing processes break down at scale. We present new software named Manubot, available at https://manubot.org , to address the challenges of open scholarly writing. Manubot adopts the contribution workflow used by many large-scale open source software projects to enable collaborative authoring of scholarly manuscripts. With Manubot, manuscripts are written in Markdown and stored in a Git repository to precisely track changes over time. By hosting manuscript repositories publicly, such as on GitHub, multiple authors can simultaneously propose and review changes. A cloud service automatically evaluates proposed changes to catch errors. Publication with Manubot is continuous: When a manuscript’s source changes, the rendered outputs are rebuilt and republished to a web page. Manubot automates bibliographic tasks by implementing citation by identifier, where users cite persistent identifiers (e.g. DOIs, PubMed IDs, ISBNs, URLs), whose metadata is then retrieved and converted to a user-specified style. Manubot modernizes publishing to align with the ideals of open science by making it transparent, reproducible, immediate, versioned, collaborative, and free of charge.

BackgroundUnsupervised compression algorithms applied to gene expression data extract latent, or hidden, signals representing technical and biological sources of variation. However, these algorithms require a user to select a biologically-appropriate latent dimensionality. In practice, most researchers select a single algorithm and latent dimensionality. We sought to determine the extent by which using multiple dimensionalities across ensemble compression models improves biological representations.ResultsWe compressed gene expression data from three large datasets consisting of adult normal tissue, adult cancer tissue, and pediatric cancer tissue. We compressed these data into many latent dimensionalities ranging from 2 to 200. We observed various tradeoffs across latent dimensionalities and compression models. For example, we observed high model stability between principal components analysis (PCA), independent components analysis (ICA), and non-negative matrix factorization (NMF). We identified more unique biological signatures in ensembles of denoising autoencoder (DAE) and variational autoencoder (VAE) models in intermediate latent dimensionalities. However, we captured the most pathway-associated features using all compressed features across algorithms and dimensionalities. Optimized at different latent dimensionalities, compression models detect generalizable gene expression signatures representing sex, neuroblastoma MYCN amplification, and cell types. In two supervised machine learning tasks, compressed features optimized predictions at different latent dimensionalities.ConclusionsThere is no single best latent dimensionality or compression algorithm for analyzing gene expression data. Instead, using feature ensembles from different compression models across latent space dimensionalities optimizes biological representations.