Giotto 1.0.1 - 1.0.3¶
Added seed to HMRF
Created functions to read 10X Visium .h5 files
see createGiottoVisiumObject to create a Giotto object directly
see get10Xmatrix_h5 to extract the count matrix
This is the first major release of Giotto.
If you still want to work with the previous version, then you can find the older releases here.
Here is an overview about what has changed in the meantime:
NEW: Addition of getSpatialDataset to directly download a spatial dataset (expression matrix, spatial coordinates and metadata). This is now also included in the examples that you can find under the Datasets tab on this website.
NEW: We have added tools to install, remove and check a Giotto r-miniconda environment. This miniconda environment is one way to make sure that you can run functions that require Python modules.
installGiottoEnvironment: (re-)installs a Giotto miniconda environment
removeGiottoEnvironment: removes a Giotto miniconda environment
checkGiottoEnvironment: verifies if a Giotto environment can be found
The other alternative is to install them in your own favorite Python environment and provide the path in the createGiottoInstructions command.
Extension and improvement of spatial gene detection methods:
NEW: addition of spark method
Improvements for silhouetteRank:
Multi parameter version as silhouetteRankTest
Improvements for binSpect:
Multi parameter version: binSpectSingle or binSpectMulti
Spatial cell type enrichment methods have been streamlined and updated
runPAGEEnrich to run enrichment using PAGE algorithm and selected marker genes
runRankEnrich to run enrichment using a whole expression matrix
runHyperGeometricEnrich to run enrichment using the hypergeometric test
NEW: Spatial cell type deconvolution has been added:
use runSpatialDeconv or runDWLSDeconv
NEW: Addition of addCellIntMetadata to add information about interacting cell types, which can subsequently be viewed with the spatPlot commands.
NEW: Addition of 3 small vignettes that cover different types of spatial datasets:
Cell Proximity Genes has been changed to Interaction Changed Genes
This better reflects the nature of gene changes due to neighboring cell interactions
CPG functions are deprecated and will be removed in the future
Several function help pages have been updated with dummy example code
several small and big fixes to the code
background images See HowTos for more information!
support for sparse matrices
PCA can be calculated with the packages irlba (default) or factominer (old default)
complemented PCA with separate functions for a scree plot and jackstraw plot
addition of readExprMatrix to read an expression matrix
addition of addGenesPerc to add information about genesets (e.g. mitochondrial genes)
addition of showGrids and showNetworks to see available spatial grids and networks
several bug fixes
added voronoi plots to use in spatial plotting. See HowTos for more informaiton.
generalized visualization parameters between functions
(optional) automatic installation of python modules through reticulate:
you can provide your preferred python path
the giotto environment can be installed automatic
if you do not provide the python path and do not choose to install the giotto environment, then it will take the default python path
several bug fixes
several mini-datasets are now included within Giotto for quick testing:
field 1 of seqFISH+ (single-cell)
the visium brain Dentate Gyrus subset (spots)
subset of starMAP (3D)
example to acces the seqFISH+ mini dataset:
# raw counts small_seqfish_expr_matrix = read.table(system.file("extdata", "seqfish_field_expr.txt", package = 'Giotto')) # cell locations small_seqfish_locations = read.table(system.file("extdata", "seqfish_field_locs.txt", package = 'Giotto'))
Default spatial network created with createSpatialNetwork is now a Delaunay spatial network.
# to create the old default kNN spatial network use: createSpatialKNNnetwork(gobject) # or use this function with the following setting createSpatialNetwork(gobject, method = 'kNN')
The function names for extracting spatial genes have changed:
# binGetSpatialGenes is now: binSpect(gobject) # binary Spatial extraction # spatial_genes_python is now: silhouetteRank(gobject)
Fixed multiple bugs
Improved speed by changing code to Rcpp and implementing parallelization options
updated HowTos tutorials in Start section
Finished the analysis of 10 different spatial datasets (tutorials are a work-in-progress)
New examples on mouse kidney and brain using the recently released 10X Visium datasets (NEW)
Added tools to identify spatial enrichment based on cell-type specific gene signature lists (NEW)
New example with 3D-like spatial data of the mouse hypothalamic preoptic region using merFISH (NEW)
New example with 3D spatial data STARmap
New example with the highly sensitive data from osmFISH
New example on the Cerebellum with the scalable data from Slideseq
New example on mouse olfactory bulb using immobilized primers on glass slides from Spatial Transcriptomics
Updated seqFISH+ cortex example (NEW)
Updated STARmap cortex example (NEW)