Abstract
Single-cell technologies offer an unprecedented opportunity to effectively characterize cellular heterogeneity in health and disease. Nevertheless, visualisation and interpretation of these multi-dimensional datasets remains a challenge. We present a novel framework, ivis, for dimensionality reduction of single-cell expression data. ivis utilizes a siamese neural network architecture that is trained using a novel triplet loss function. Results on simulated and real datasets demonstrate that ivis preserves global data structures in a low-dimensional space, adds new data points to existing embeddings using a parametric mapping function, and scales linearly to hundreds of thousands of cells. ivis is made publicly available through Python and R interfaces on https://github.com/beringresearch/ivis.
Highlights
Characterising cellular composition is crucial for defining functional heterogeneity in health and disease[1]
Visualisation and interpretation of single-cell experiments are underpinned by dimensionality reduction (DR) techniques
Unsupervised Neural Network (NN) with multiple layers are trained by optimizing a target function, whilst an intermediate layer with small cardinality serves as a low dimensional representation of the input data[19,21]
Summary
Characterising cellular composition is crucial for defining functional heterogeneity in health and disease[1]. Non-linear approaches, including the t-distributed Stochastic Neighbor Embedding (t-SNE) algorithm[11], have been shown to effectively capture complex data structures, outperforming linear projection methods such as Principal Component Analysis (PCA)[12,13] t-SNE has several limitations[14,15]. Due to non-parametric nature of t-SNE, addition of new data points to existing embeddings is not possible[11,15]. In this paper we introduce a scalable algorithm, ivis, which effectively captures local as well as global features of high-dimensional datasets. Ivis learns a parametric mapping from the high-dimensional space to low-dimensional embedding, facilitating seamless addition of new data points to the mapping function. We demonstrate that ivis preserves distances in low-dimensional projections, enabling biological interpretation. We validate our method using synthetic, cytometry by time of flight (CyTOF), and scRNA-seq datasets
Sign-in/Register to view highlights & summary 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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
