- Nov 5, 2024
- update to Bioc 3.20
- Oct 27, 2024
- Better error reporting addressing #25
- July 19, 2024
- fix in
.read_matrix_block()
- See https://github.com/GabrielHoffman/dreamlet/pull/23/commits/c453ac98ebc0329279b4dd3ae26a674df0e9b1f2
- May 31, 2024
- bump Bioc version
- May 28, 2024
dreamlet()
gives clearer error message for singular design matrixplotGeneHeatmap()
handleszmax
correctly now
- May 16, 2024
processOneAssay
, setrescaleWeightsAfter=FALSE
by default
- May 14, 2024
- fix issue in
topTable()
with multiplecoef
values
- fix issue in
run_mash()
with multiple coefficients- from #18
- in
dreamlet()
allow formula to include only intercept
- bug fix in
compositePosteriorTest()
- bug fix
- new functionality in
plotPCA()
andoutlierByAssay()
- works on any
list
, not justdreamletProcessedData
- allows outlier analysis on residuals
- Feb 8, 2024
- fix bug in
pbWeights()
- Fix scaling issue in
outlierByAssay()
- Feb 5, 2024
- Fix bug in call to
eBayes()
- in
processAssays()
pass argumentscaledByLib
tovoomWithDreamWeights()
- Jan 29, 2024
- fix bug in
pbWeights()
- smaller pseudo variance
- limit to only expressed genes by adding
getExprGeneNames()
- Jan 25, 2024
- improve error reporting in
seeErrors()
and documentation - update
outlier()
to compute z-scores. How returnsdata.frame()
- add
outlierByAssay()
andplotPCA()
- Jan 16, 2024
compositePosteriorTest()
allowsexclude
set to beNULL
- Jan 16, 2024
- add
meta_analysis()
- Jan 10, 2024
stackAssays()
now includedsmetadata()$aggr_means
correctly- add
compositePosteriorTest()
- Jan 3, 2024
- use
get_metadata_aggr_means()
to extractaggr_means
when SCE is produced by cbind'ing
- Dec 18, 2023
- fix issue when no genes pass cutoffs
- fix issue with
aggr_means
inaggregateToPseudoBulk()
- fix bug when
rdf
is low for all genes
- Dec 10, 2023
- add
plotBeeswarm()
- add
rowWeightedVarsMatrix()
- bug fixes
- add
isFullRank()
check indreamlet()
- handle exceptions in
run_mash()
- Dec 5, 2023
pbWeights()
add argumentmaxRatio
- Nov 28, 2023
- fix edge cases in
pbWeights()
- cell weights is not default in
dreamlet()
- Nov 27, 2023
- add
pbWeights()
to compute precision weights for pseudobulk counts - extend
extractData()
and include it in vignette
- Nov 22, 2023
- add
stackAssays()
- add
diffVar()
- fix
getVarFromCounts)
sozeta
is a mean, not a sum
- Nov 13, 2023
computeLogCPM()
usesaugmentPriorCount()
- bug fix in Bioc 3.18 and devel
computeLogCPM()
now returnsmatrix
instead ofsparseMatrix
- Oct 18, 2023
- update
variancePartition
version dependency - add
getWeightsList()
- Oct 12, 2023
- fix error reporting in
processAssays()
- Sept 22, 2023
- Fix issue with precision weights from cell number
- Include
setAutoBlockSize()
update withinaggregateToPseudoBulk()
- Sept 5, 2023
- Update error handling for
processAssays()
andfitVarPart()
- August 18, 2023
- Update error handling and documentation
- August 8, 2023
- run
styler::style_pkg()
- August 8, 2023
dreamletCompareClusters()
now allows cell-level covariates in response to #11- Fix code for Bioconductor submission
- July 17, 2023
- Improve functionality and documentation of
dreamlet::residuals()
- June 29, 2023
- in
processAssays()
usevoomWithDreamWeights(..., span="auto")
to estimate the lowess tuning parameter - rare error in
merge_metadata()
when a cell type is not observed for all donors.
- June 28, 2023
- in
dreamlet()
fix issue when contrasts are specified and formula includes variable frommetadata()
- June 27, 2023
- add
assays
argument tobuildClusterTreeFromPB()
- June 16, 2023
- bug fix in
processAssays()
when assays is dropped
- May 31, 2023
- improved error reporting
- Compatibility with variancePartition v2.0.5 (renamed 1.31.1)
- May 24, 2023
- required
zellkonverter (>= 1.10.1)
to avoid issues with previous version
- May 12, 2023
- Compatibility with variancePartition v2.0.1
- April 24, 2023
- Fix issue in raised in Bioconductor submission: Bioconductor/Contributions#2955 (comment)
- Compatibility with variancePartition v2.0.0
- March 29, 2023
- fix
topTable()
fordreamletResult
in the case where one or more cells didn't estimate the coefficient of interest
- March 23, 2023
- reduce compiler time
- add
computeNormCounts()
andcomputeLogCPM()
- March 20, 2023
- fix Biocondcutor submission based on Bioconductor/Contributions#2955 (comment)
- March 15, 2023
- Biocondcutor submission
- March 7, 2023
- fix
topTable()
to deal with multiplecoef
as array - add vignette about non-linear effects
- Add
colData<-
fordreamletProcessedData
- March 3, 2023
- small bug fixes in
topTable()
andplotForest()
- Feb 27, 2023
aggregateToPseudoBulk()
stores mean of cell-level covariates inmetadata(pb)$aggr_means
- Use these covariates in
processAssays()
,dreamlet()
,fitVarPart()
- extend to
aggregateNonCountSignal()
- Jan 25, 2023
- add
plotProjection()
- add
outlier()
- update
plotForest()
- Jan 18, 2023
- allow custom ordering of assays in plots
- Jan 12, 2023
- update
plotVolcano()
to allowscales="free_y"
- warning when p-values are zero.
- Jan 9, 2023
- Rewrite
aggregateNonCountSignal()
to include filters - Depend only on CRAN and BioC packages
- Jan 4, 2023
- Include precision weights in
aggregateNonCountSignal()
- Jan 3, 2023
- Enable processing of non-count data with
aggregateNonCountSignal()
- Dec 7, 2022
- in
plotGeneHeatmap()
drop empty genes
- Nov 30, 2022
- add
buildClusterTreeFromPB()
- Nov 18, 2022
- fix bug in
topTable()
- Nov 18, 2022
- add
as.dreamletResult()
- update
variancePartition
dependency and source
- Nov 10, 2022
- in
processAssays()
andprocessOneAssay()
, add argumentmin.prop
indicating the minimum proportion of retained samples with non-zero counts
- Nov 2, 2022
- fix bug in
zenith_gsa()
for few gene sets
- Oct 17, 2022
- add
computeCellCounts()
- Sept 14, 2022
- add
transpose
argument toplotGeneHeatmap()
- and
alpha
arugment toplotVoom()
- update y axis and outlines in
plotVarPart()
- update filtering of covariates, especially for when many samples are dropped
- add
totalCPM
column to output ofcellTypeSpecificity()
to use for filtering. Functionsdreamlet::plotHeatmap()
plotViolin()
andplotPercentBars()
now ignore this column
- update documentation
- add
plotGeneHeatmap()
- add argument
assays
toplotVarPart()
- add
extractData()
- faster
aggregateToPseudoBulk()
by speeding up check in.check_arg_assay()
- more flexibility for
tabToMatrix()
- fix misc issues with plotting and order of facets
- fix issue with redundant variables in small sample sizes
- fix issue with
topTable()
when all random effects are dropped
- Compatibility with zenith package submitted to Bioconductor.
- fix issue with
aggregateToPseudoBulk()
when summarizing for just 1 sample
- add
getTreat()
fordreamlet()
result
droplevels
forcolData
inprocessAssays()
- fixes in
processAssays()
to detect issues with SCE
- remove old code and dependencies
- change website theme
- bug fix for
aggregateToPseudoBulk()
with sample ordering
- bug fix for
dreamletCompareClusters()
- add
colsum2()
using beachmat code.
- reduce memory usage in
aggregateToPseudoBulk()
by fixing `aggregateByColnames()
- update
run_mash()
to combine results across coefs
- fix bug in
dreamlet::colsum_fast()
used in pseudobulk
- add
da_to_sparseMatrix()
aggregateToPseudoBulk()
forDelayedArray
now usescolsum_fast()
- this is faster then the previous version for
DelayedArray
- this is faster then the previous version for
- update
dreamletCompareClusters()
:- now compatable with
plotZenithResults()
- include flag
errorsAsWarnings
. IfTRUE
warns and returns NULL.
- now compatable with
- change return value for
dreamletCompareClusters()
to be compatible withzenith_gsa()
- fix usage of
formula
indreamletCompareClusters()
- additional speed up
aggregateToPseudoBulk()
when a Seurat object is used- uses RcppEigen sparse matrix iterator
- dramatic speed up
aggregateToPseudoBulk()
when a Seurat object is used- uses RcppEigen
- Speed up
aggregateToPseudoBulk()
when a Seurat object is used
- Add
collapse=TRUE
todreamletCompareClusters()
- Fix bug in
dreamletCompareClusters()
- Fix bug in
dreamletCompareClusters()
- add
min.samples
toprocessAssays()
,processOneAssay()
- add
dreamletCompareClusters()
andrun_mash()
- Fix bug in
dreamletCompareClusters()
- updated
mashr
dependency
- add
run_mash()
- add
zenith_gsa()
,plotVolcano()
,plotForest()
for results
- add
- fix bug in
cellTypeSpecificity()
for genes with zero reads across all cell types - order of arguments in
plotForest()
andzenith_gsa()
changed for consistancy - expand vignettes
- bug fix for
removeConstantTerms()
when excluded variable string (i.e. tissue) is also a substring of other variables (i.e. tissueStatus)
- add
residuals()
fordreamlet()
result - add
dreamletPairs()
- fix bug in
removeConstantTerms()
with multiple constant terms - improve usability of
cellTypeSpecificity()
by addingplotPercentBars()
andplotViolin()
compatability - fix bug in
topTable()
whencoef
is not estimated - add argument
assays
todreamlet()
,fitVarPart()
, andprocessAssays()
processOneAssay()
weights by number of cells- require
variancePartition >= 1.25.1
to handle weights invoomWithDreamWeights()
- fix bug in
topTable()
- add
plotPercentBars()
for classvpDF
- misc bug fixes
- improve documentation
- move count ratio code to crumblr package
- use
applyQualityWeights()
- Oct 25, 2021
- add
ilr_composition_test.R
- Oct 15, 2021
- update print for
dreamletResult
usingcoefNames()
- small bug fix
- fix bugs in
regModel()
- Oct 6, 2021
removeConstantTerms()
now doesn't drop terms solely because of NA's- this means that other functions can gracefully warn the user about NA's
-
Oct 5, 2021
-
suppress package startup messages in
aggregateToPseudoBulk()
-
bug fix in
removeConstantTerms()
-
Sept 30, 2021
-
call to
zenith_gsa()
adds argumentinter.gene.cor
andprogressbar
-
fix output to
cellTypeCompositionVarPart()
andcellTypeCompositionTest()
-
fix issue with
topTable()
where FDR was evaluated on only a subset of genes
- Sept 28, 2021
- update docs, logos, TODO
- update
dreamlet()
to handle linear contrasts removeConstantTerms()
now drops categorical variables with only a max of one example per category- Cleaner code for cell composition test
- Sept 2, 2021
- add
cellTypeCompositionTest()
- handle random effects
- Sept 1, 2021
- enforce package version requirements
- August 25, 2021
- handling of pmetadata by processAssays(), fitVarPart(), and dreamlet()
- change defaults for bpparam to SerialParam()
- Created new files for code
- Create object dreamletResult returned by dreamlet() and used by topTable()
- more capable dreamletProcessedData object
- add details() functions for dreamletResult, dreamletProcessedData and vpDF
- More information about when terms are dropped from formulas
- type definition to zenith_gsa
- add aggregateToPseudoBulk() for faster access to SingleCellExperiment() backed by H5AD
- add argument to
processAssays()
to include extra meta-data - add subseting with assay() for dreamletProcessedData
- adapt plotVoom(), plotVolcano(), plotVarPart() to be more flexiable
- fitVarPart() returns DataFrame
- check failed regression
- Initial version