Abstract
MotivationPseudotime estimation from single-cell gene expression data allows the recovery of temporal information from otherwise static profiles of individual cells. Conventional pseudotime inference methods emphasize an unsupervised transcriptome-wide approach and use retrospective analysis to evaluate the behaviour of individual genes. However, the resulting trajectories can only be understood in terms of abstract geometric structures and not in terms of interpretable models of gene behaviour.ResultsHere we introduce an orthogonal Bayesian approach termed ‘Ouija’ that learns pseudotimes from a small set of marker genes that might ordinarily be used to retrospectively confirm the accuracy of unsupervised pseudotime algorithms. Crucially, we model these genes in terms of switch-like or transient behaviour along the trajectory, allowing us to understand why the pseudotimes have been inferred and learn informative parameters about the behaviour of each gene. Since each gene is associated with a switch or peak time the genes are effectively ordered along with the cells, allowing each part of the trajectory to be understood in terms of the behaviour of certain genes. We demonstrate that this small panel of marker genes can recover pseudotimes that are consistent with those obtained using the entire transcriptome. Furthermore, we show that our method can detect differences in the regulation timings between two genes and identify ‘metastable’ states—discrete cell types along the continuous trajectories—that recapitulate known cell types.Availability and implementationAn open source implementation is available as an R package at http://www.github.com/kieranrcampbell/ouija and as a Python/TensorFlow package at http://www.github.com/kieranrcampbell/ouijaflow.Supplementary information Supplementary data are available at Bioinformatics online.
Highlights
We demonstrate that this small panel of marker genes can recover pseudotimes that are consistent with those obtained using the entire transcriptome
4.1 Pseudotime inference from small marker gene panels. The transcriptomes of both single cells and bulk samples exhibit remarkable correlations across genes and transcripts. Such concerted regulation of expression is thought to be due to pathway-dependent transcription (Tegge et al, 2012; Braun et al, 2008) and is necessary for the field of network inference from gene expression data (Langfelder and Horvath, 2008)
We further investigated the single cell expression data from a study tracking the differentiation of embryonic precursor cells into haematopoietic stem cells (HSCs) (Zhou et al, 2016)
Summary
In this paper we present an orthogonal approach implemented within a Bayesian latent variable statistical framework called ‘Ouija’ that learns pseudotimes from small panels of putative or known marker genes (Figure 1A). Our model focuses on switch-like and transient expression behaviour along pseudotime trajectories, explicitly modelling when a gene turns on or off along a trajectory or at which point its expression peaks. This allows the pseudotime inference procedure to be understood in terms of descriptive gene regulation events along the trajectory (Figure 1B). We further formulate a Bayesian hypothesis test as to whether a given gene is regulated before another along the pseudotemporal trajectory (Figure 1C) for all pairwise combinations of genes By using such a probabilistic model we can identify discrete cell types or “metastable states” along continuous developmental trajectories (Figure 1D) that correspond to known cell types
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
