Abstract:MOTIVATION:Single-cell technologies have recently gained popularity in cellular differentiation studies regarding their ability to resolve potential heterogeneities in cell populations. Analyzing such high-dimensional single-cell data has its own statistical and computational challenges. Popular multivariate approaches are based on data normalization, followed by dimension reduction and clustering to identify subgroups. However, in the case of cellular differentiation, we would not expect clear clusters to be present but instead expect the cells to follow continuous branching lineages.RESULTS:Here, we propose the use of diffusion maps to deal with the problem of defining differentiation trajectories. We adapt this method to single-cell data by adequate choice of kernel width and inclusion of uncertainties or missing measurement values, which enables the establishment of a pseudotemporal ordering of single cells in a high-dimensional gene expression space. We expect this output to reflect cell differentiation trajectories, where the data originates from intrinsic diffusion-like dynamics. Starting from a pluripotent stage, cells move smoothly within the transcriptional landscape towards more differentiated states with some stochasticity along their path. We demonstrate the robustness of our method with respect to extrinsic noise (e.g. measurement noise) and sampling density heterogeneities on simulated toy data as well as two single-cell quantitative polymerase chain reaction datasets (i.e. mouse haematopoietic stem cells and mouse embryonic stem cells) and an RNA-Seq data of human pre-implantation embryos. We show that diffusion maps perform considerably better than Principal Component Analysis and are advantageous over other techniques for non-linear dimension reduction such as t-distributed Stochastic Neighbour Embedding for preserving the global structures and pseudotemporal ordering of cells.AVAILABILITY AND IMPLEMENTATION:The Matlab implementation of diffusion maps for single-cell data is available at https://www.helmholtz-muenchen.de/icb/single-cell-diffusion-map.CONTACT:fbuettner.phys@gmail.com, fabian.theis@helmholtz-muenchen.deSUPPLEMENTARY INFORMATION:Supplementary data are available at Bioinformatics online.

Dimensional reduction of gradient-like stochastic systems with multiplicative noise via Fokker-Planck diffusion maps

Data-Driven Stochastic Averaging

Data-driven Method for Dimension Reduction of Nonlinear Randomly Vibrating Systems

Data-driven Dimension Reduction for High-Dimensional Random Vibration Systems with Cubic Nonlinearity

Diffusion Maps for High-Dimensional Single-Cell Analysis of Differentiation Data

Dimensionality Reduction in Stochastic Complex Dynamical Networks

Dimensionality Reduction for Classification of Stochastic Texture Images

Large-Scale Wasserstein Gradient Flows

Dimension reduction for large-scale stochastic systems with non-zero initial states and controlled diffusion

Statistical algorithms for low-frequency diffusion data: A PDE approach

Diffusion Maps Kalman Filter for a Class of Systems With Gradient Flows

Potential Landscape of High Dimensional Nonlinear Stochastic Dynamics with Large Noise

Dimension reduction and the gradient flow of relative entropy

ADM: Adaptive Graph Diffusion for Meta-Dimension Reduction

Improved visualization of high-dimensional data using the distance-of-distance transformation

Fluctuations in Wasserstein dynamics on Graphs

Identifying Drift, Diffusion, and Causal Structure from Temporal Snapshots

Potential landscape of high dimensional nonlinear stochastic dynamics and rare transitions with large noise

Beating the curse of dimension with accurate statistics for the Fokker–Planck equation in complex turbulent systems

Deterministic Fokker-Planck Transport -- With Applications to Sampling, Variational Inference, Kernel Mean Embeddings & Sequential Monte Carlo

Adapting to Unknown Low-Dimensional Structures in Score-Based Diffusion Models