Moved the Global score summaries of SpatialBivariate from .uns to .var
df_to_lr will now also return the expression and proportion of expression for the interactions
li.multi.nfm will now also accept a DataFrame as input
Filtered putative interactions in the Consensus resource, coming from CellTalkDB.
Changed filter_lambda parameter to filter_fun for consistency and now any function can be passed to be applied as a row-wise filter.
Global results of SpatialBivariate will now be saved to .var
LIANA's spatial methods will now work with non-aligned AnnData objects, i.e. when observations across modalities are not aligned.
= Added li.ut.interpolate_adata utility function to interpolate the data to a common space.
= MISTy will also work with directly non-aligned data with spatial connectivities from one modality to the other being passed via obsm rather than obsp. Making use of li.ut.spatial_neighbors by passing reference coordinates.
Fixed a bug where li.ut.obsm_to_adata would assign var as a method rather than DataFrame
Fixed a bug where p-values for Global Moran's were not calculated correctly.
Enabled cell_pairs of interest to be passed to single-cell methods.
Enabled Parallelization of Permutation-based methods.
Local categories will now be only calculated for positive interactions (not non-ambigous as before).
Names of source and target can now be passed to li.pl.tileplot.
1.0.4 (02.01.2024)
Moved the Global score summaries of
SpatialBivariatefrom .uns to .vardf_to_lrwill now also return the expression and proportion of expression for the interactionsli.multi.nfmwill now also accept a DataFrame as inputFiltered putative interactions in the Consensus resource, coming from CellTalkDB.
Changed
filter_lambdaparameter tofilter_funfor consistency and now any function can be passed to be applied as a row-wise filter.Global results of
SpatialBivariatewill now be saved to.varLIANA's spatial methods will now work with non-aligned AnnData objects, i.e. when observations across modalities are not aligned. = Added
li.ut.interpolate_adatautility function to interpolate the data to a common space. = MISTy will also work with directly non-aligned data with spatial connectivities from one modality to the other being passed viaobsmrather thanobsp. Making use ofli.ut.spatial_neighborsby passing reference coordinates.Fixed a bug where
li.ut.obsm_to_adatawould assign var as a method rather than DataFrameFixed a bug where p-values for Global Moran's were not calculated correctly.
Enabled
cell_pairsof interest to be passed to single-cell methods.Enabled Parallelization of Permutation-based methods.
Local categories will now be only calculated for positive interactions (not non-ambigous as before).
Names of source and target can now be passed to
li.pl.tileplot.