Find open-source science resources

Useful functions to visualize single cell and spatial data. It supports visualizing 'Seurat', 'SingleCellExperiment' and 'SpatialExperiment' objects through grammar of graphics syntax implemented in 'ggplot2'.

Provides an R wrapper for the implementation of FI-tSNE from the python package openTNSE. See Poličar et al. (2019) <doi:10.1101/731877> and the algorithm described by Linderman et al. (2018) <doi:10.1038/s41592-018-0308-4>.

scTGIF connects the cells and the related gene functions without cell type label.

The algorithm is based on the non-negative tucker decomposition (NTD2) of nnTensor.

Many modern biological datasets consist of small counts that are not well fit by standard linear-Gaussian methods such as principal component analysis. This package provides implementations of count-based feature selection and dimension reduction algorithms. These methods can be used to facilitate unsupervised analysis of any high-dimensional data such as single-cell RNA-seq.

A package for nonlinear dimension reduction using the Isomap and LLE algorithm. It also includes a routine for computing the Davis-Bouldin-Index for cluster validation, a plotting tool and a data generator for microarray gene expression data and for the Swiss Roll dataset.

This package provides a comprehensive set of external and internal evaluation metrics. It includes metrics for assessing partitions or fuzzy partitions derived from clustering results, as well as for evaluating subpopulation identification results within embeddings or graph representations. Additionally, it provides metrics for comparing spatial domain detection results against ground truth labels, and tools for visualizing spatial errors.

The MOFA2 package contains a collection of tools for training and analysing multi-omic factor analysis (MOFA). MOFA is a probabilistic factor model that aims to identify principal axes of variation from data sets that can comprise multiple omic layers and/or groups of samples. Additional time or space information on the samples can be incorporated using the MEFISTO framework, which is part of MOFA2. Downstream analysis functions to inspect molecular features underlying each factor, visualisation, imputation etc are available.

F-informed MDS is a new multidimensional scaling-based ordination method that configures data distribution based on the F-statistic (i.e., the ratio of dispersion between groups with shared or differing labels).

Implements the density-preserving modification to t-SNE and UMAP described by Narayan et al. (2020) <doi:10.1101/2020.05.12.077776>. The non-linear dimensionality reduction techniques t-SNE and UMAP enable users to summarise complex high-dimensional sequencing data such as single cell RNAseq using lower dimensional representations. These lower dimensional representations enable the visualisation of discrete transcriptional states, as well as continuous trajectory (for example, in early development). However, these methods focus on the local neighbourhood structure of the data. In some cases, this results in misleading visualisations, where the density of cells in the low-dimensional embedding does not represent the transcriptional heterogeneity of data in the original high-dimensional space. den-SNE and densMAP aim to enable more accurate visual interpretation of high-dimensional datasets by producing lower-dimensional embeddings that accurately represent the heterogeneity of the original high-dimensional space, enabling the identification of homogeneous and heterogeneous cell states. This accuracy is accomplished by including in the optimisation process a term which considers the local density of points in the original high-dimensional space. This can help to create visualisations that are more representative of heterogeneity in the original high-dimensional space.