New in 0.16.0 (2022-04-12)#
This release features a refactor of DestVI
(#1457):
Bug fix in cell type amortization, which leads to on par performance of cell type amortization
V_encoder
with free parameter for cell type proportionsV
.Bug fix in library size in
CondSCVI
, that lead to downstream dependency between sum over cell type proportionsv_ind
and library sizelibrary
inDestVI
.neg_log_likelihood_prior
is not computed anymore on random subset of single cells but cell type specific subclustering using cluster variancevar_vprior
, cluster meanmean_vprior
and cluster mixture proportionmp_vprior
for computation. This leads to more stable results and faster computation time. Settingvamp_prior_p
infrom_rna_model()
to the expected resolution is critical in this algorithm.The new default is to also use dropout
dropout
during the decoder ofCondSCVI
and subsequentlydropout_decoder
inDestVI
, we found this to be beneficial after bug fixes listed above.We changed the weighting of the loss on the variances of beta and the prior of eta.
Note
Due to bug fixes listed above this version of DestVI
is not backwards compatible. Despite instability in training in the
outdated version, we were able to reproduce results generated with this code. We therefore do not strictly encourage to rerun old experiments.
We published a new tutorial. This new tutorial incorporates a new utility package destvi_utils that generates exploratory plots of the results of DestVI
. We refer to the manual of this package for further documentation.
Changes#
Breaking changes#
Support for loading legacy loading is removed from
load()
. Utility to convert old files to the new file as been addedconvert_legacy_save()
(#1505).
Bug Fixes#
get_likelihood_parameters()
fix forn_samples > 1
anddispersion="gene_cell"
#1504.Fix backwards compatibility for legacy TOTALVI models #1502.