Closed FerrenaAlexander closed 1 year ago
I'm having the same problem.
what is the output of rsconnect::appDependencies() ? Thanks!
here is my output...I'm trying to publish single-cell-viewer to shinyapps.io
> options(repos = BiocManager::repositories())
> rsconnect::appDependencies()
package version source
1 BH 1.72.0-3 CRAN
2 Biobase 2.46.0 Bioconductor
3 BiocGenerics 0.32.0 Bioconductor
4 DT 0.11 CRAN
5 FNN 1.1.3 CRAN
6 KernSmooth 2.23-16 CRAN
7 MASS 7.3-51.5 CRAN
8 Matrix 1.2-18 CRAN
9 R.methodsS3 1.7.1 CRAN
10 R.oo 1.23.0 CRAN
11 R.utils 2.9.2 CRAN
12 R6 2.4.1 CRAN
13 RANN 2.6.1 CRAN
14 RColorBrewer 1.1-2 CRAN
15 ROCR 1.0-7 CRAN
16 RSpectra 0.16-0 CRAN
17 Rcpp 1.0.3 CRAN
18 RcppAnnoy 0.0.14 CRAN
19 RcppEigen 0.3.3.7.0 CRAN
20 RcppParallel 4.4.4 CRAN
21 RcppProgress 0.4.1 CRAN
22 Rdpack 0.11-1 CRAN
23 Rtsne 0.15 CRAN
24 SDMTools 1.1-221.2 CRAN
25 Seurat 3.1.2 Bioconductor
26 TFisher 0.2.0 CRAN
27 TH.data 1.0-10 CRAN
28 ape 5.3 CRAN
29 askpass 1.1 CRAN
30 assertthat 0.2.1 CRAN
31 backports 1.1.5 CRAN
32 base64enc 0.1-3 CRAN
33 bibtex 0.4.2.2 CRAN
34 bitops 1.0-6 CRAN
35 caTools 1.17.1.4 CRAN
36 canvasXpress 1.26.5 CRAN
37 cli 2.0.1 CRAN
38 cluster 2.1.0 CRAN
39 codetools 0.2-16 CRAN
40 colorspace 1.4-1 CRAN
41 cowplot 1.0.0 CRAN
42 crayon 1.3.4 CRAN
43 crosstalk 1.0.0 CRAN
44 curl 4.3 CRAN
45 data.table 1.12.8 CRAN
46 digest 0.6.23 CRAN
47 dplyr 0.8.3 CRAN
48 dqrng 0.2.1 CRAN
49 ellipsis 0.3.0 CRAN
50 fansi 0.4.1 CRAN
51 farver 2.0.1 CRAN
52 fastmap 1.0.1 CRAN
53 fitdistrplus 1.0-14 CRAN
54 future 1.15.1 CRAN
55 future.apply 1.4.0 CRAN
56 gbRd 0.4-11 CRAN
57 gdata 2.18.0 CRAN
58 ggplot2 3.2.1 CRAN
59 ggrepel 0.8.1 CRAN
60 ggridges 0.5.2 CRAN
61 globals 0.12.5 CRAN
62 glue 1.3.1 CRAN
63 gplots 3.0.1.2 CRAN
64 gridExtra 2.3 CRAN
65 gtable 0.3.0 CRAN
66 gtools 3.8.1 CRAN
67 hexbin 1.28.0 CRAN
68 htmltools 0.4.0 CRAN
69 htmlwidgets 1.5.1 CRAN
70 httpuv 1.5.2 CRAN
71 httr 1.4.1 CRAN
72 ica 1.0-2 CRAN
73 igraph 1.2.4.2 CRAN
74 irlba 2.3.3 CRAN
75 jsonlite 1.6 CRAN
76 labeling 0.3 CRAN
77 later 1.0.0 CRAN
78 lattice 0.20-38 CRAN
79 lazyeval 0.2.2 CRAN
80 leiden 0.3.1 CRAN
81 lifecycle 0.1.0 CRAN
82 listenv 0.8.0 CRAN
83 lmtest 0.9-37 CRAN
84 logging 0.10-108 CRAN
85 lsei 1.2-0 CRAN
86 lubridate 1.7.4 CRAN
87 magrittr 1.5 CRAN
88 metap 1.2 CRAN
89 mgcv 1.8-31 CRAN
90 mime 0.8 CRAN
91 mnormt 1.5-5 CRAN
92 multcomp 1.4-12 CRAN
93 multtest 2.42.0 Bioconductor
94 munsell 0.5.0 CRAN
95 mutoss 0.1-12 CRAN
96 mvtnorm 1.0-12 CRAN
97 nlme 3.1-143 CRAN
98 npsurv 0.4-0 CRAN
99 numDeriv 2016.8-1.1 CRAN
100 openssl 1.4.1 CRAN
101 openxlsx 4.1.4 CRAN
102 pbapply 1.4-2 CRAN
103 periscope 0.4.7 CRAN
104 pillar 1.4.3 CRAN
105 pkgconfig 2.0.3 CRAN
106 plogr 0.2.0 CRAN
107 plotly 4.9.1 CRAN
108 plotrix 3.7-7 CRAN
109 plyr 1.8.5 CRAN
110 png 0.1-7 CRAN
111 promises 1.1.0 CRAN
112 purrr 0.3.3 CRAN
113 reshape2 1.4.3 CRAN
114 reticulate 1.13 CRAN
115 rlang 0.4.2 CRAN
116 rsvd 1.0.2 CRAN
117 sandwich 2.5-1 CRAN
118 scales 1.1.0 CRAN
119 sctransform 0.2.1 CRAN
120 shiny 1.4.0 CRAN
121 shinyBS 0.61 CRAN
122 shinydashboard 0.7.1 CRAN
123 shinydashboardPlus 0.7.0 CRAN
124 shinyjs 1.1 CRAN
125 sitmo 2.0.1 CRAN
126 sn 1.5-4 CRAN
127 sourcetools 0.1.7 CRAN
128 stringi 1.4.5 CRAN
129 stringr 1.4.0 CRAN
130 survival 3.1-8 CRAN
131 sys 3.3 CRAN
132 tibble 2.1.3 CRAN
133 tidyr 1.0.0 CRAN
134 tidyselect 0.2.5 CRAN
135 tsne 0.1-3 CRAN
136 utf8 1.1.4 CRAN
137 uwot 0.1.5 CRAN
138 vctrs 0.2.1 CRAN
139 viridisLite 0.3.0 CRAN
140 withr 2.1.2 CRAN
141 xtable 1.8-4 CRAN
142 yaml 2.2.0 CRAN
143 zeallot 0.1.0 CRAN
144 zip 2.0.4 CRAN
145 zoo 1.8-7 CRAN
Thank you
Since your Seurat version is from Bioconductor rsconnect/packrat try to install the same version from the same repository when deploying the app. My suggestion would be to remove the package and reinstall it from cran and try again.
Interesting, not sure how that would have happened. I installed Seurat from CRAN originally.
Regardless, unininstalled Seurat and reinstalled with install.packages('Seurat')
. However, running rsconnect::appDependencies()
still shows Seurat source as Bioconductor.
Any other ideas to try? Thank you
This is what the DESCRIPTION file for Seurat looks like:
Package: Seurat
Version: 3.1.2
Date: 2019-12-12
Title: Tools for Single Cell Genomics
Description: A toolkit for quality control, analysis, and exploration of single cell RNA sequencing data. 'Seurat' aims to enable users to identify and interpret sources of heterogeneity from single cell transcriptomic measurements, and to integrate diverse types of single cell data. See Satija R, Farrell J, Gennert D, et al (2015) <doi:10.1038/nbt.3192>, Macosko E, Basu A, Satija R, et al (2015) <doi:10.1016/j.cell.2015.05.002>, and Butler A and Satija R (2017) <doi:10.1101/164889> for more details.
Authors@R: c(
person(given = 'Rahul', family = 'Satija', email = 'rsatija@nygenome.org', role = 'aut', comment = c(ORCID = '0000-0001-9448-8833')),
person(given = 'Andrew', family = 'Butler', email = 'abutler@nygenome.org', role = 'aut', comment = c(ORCID = '0000-0003-3608-0463')),
person(given = 'Paul', family = 'Hoffman', email = 'nygcSatijalab@nygenome.org', role = c('aut', 'cre'), comment = c(ORCID = '0000-0002-7693-8957')),
person(given = 'Tim', family = 'Stuart', email = 'tstuart@nygenome.org', role = 'aut', comment = c(ORCID = '0000-0002-3044-0897')),
person(given = 'Jeff', family = 'Farrell', email = 'jfarrell@g.harvard.edu', role = 'ctb'),
person(given = 'Shiwei', family = 'Zheng', email = 'szheng@nygenome.org', role = 'ctb', comment = c(ORCID = '0000-0001-6682-6743')),
person(given = 'Christoph', family = 'Hafemeister', email = 'chafemeister@nygenome.org', role = 'ctb', comment = c(ORCID = '0000-0001-6365-8254')),
person(given = 'Patrick', family = 'Roelli', email = 'proelli@nygenome.org', role = 'ctb'),
person(given = "Yuhan", family = "Hao", email = 'yhao@nygenome.org', role = 'ctb', comment = c(ORCID = '0000-0002-1810-0822'))
)
URL: http://www.satijalab.org/seurat,
https://github.com/satijalab/seurat
BugReports: https://github.com/satijalab/seurat/issues
Additional_repositories: https://mojaveazure.github.io/loomR
Depends: R (>= 3.4.0), methods,
Imports: ape, cluster, cowplot, fitdistrplus, future, future.apply,
ggplot2 (>= 3.0.0), ggrepel, ggridges, graphics, grDevices,
grid, httr, ica, igraph, irlba, KernSmooth, leiden (>= 0.3.1),
lmtest, MASS, Matrix (>= 1.2-14), metap, pbapply, plotly, png,
RANN, RColorBrewer, Rcpp, RcppAnnoy, reticulate, rlang, ROCR,
rsvd, Rtsne, scales, sctransform (>= 0.2.0), SDMTools, stats,
tools, tsne, utils, uwot (>= 0.1.5)
LinkingTo: Rcpp (>= 0.11.0), RcppEigen, RcppProgress
License: GPL-3 | file LICENSE
LazyData: true
Collate: 'RcppExports.R' 'generics.R' 'clustering.R' 'visualization.R'
'convenience.R' 'data.R' 'differential_expression.R'
'dimensional_reduction.R' 'integration.R' 'objects.R'
'preprocessing.R' 'tree.R' 'utilities.R' 'zzz.R'
RoxygenNote: 7.0.2
Encoding: UTF-8
biocViews:
Suggests: loomR, testthat, hdf5r, S4Vectors, SummarizedExperiment,
SingleCellExperiment, MAST, DESeq2, BiocGenerics,
GenomicRanges, GenomeInfoDb, IRanges, rtracklayer, monocle,
Biobase, VGAM
NeedsCompilation: yes
Packaged: 2019-12-12 21:14:37 UTC; paul
Author: Rahul Satija [aut] (<https://orcid.org/0000-0001-9448-8833>),
Andrew Butler [aut] (<https://orcid.org/0000-0003-3608-0463>),
Paul Hoffman [aut, cre] (<https://orcid.org/0000-0002-7693-8957>),
Tim Stuart [aut] (<https://orcid.org/0000-0002-3044-0897>),
Jeff Farrell [ctb],
Shiwei Zheng [ctb] (<https://orcid.org/0000-0001-6682-6743>),
Christoph Hafemeister [ctb] (<https://orcid.org/0000-0001-6365-8254>),
Patrick Roelli [ctb],
Yuhan Hao [ctb] (<https://orcid.org/0000-0002-1810-0822>)
Maintainer: Paul Hoffman <nygcSatijalab@nygenome.org>
Repository: CRAN
Date/Publication: 2019-12-12 22:20:06 UTC
In particular, because it contains a biocViews:
field, Packrat infers that the package should be restored from Bioconductor. I would suggest asking the package author to remove that field from their CRAN submission; it should only be included for packages available on Bioconductor.
@kevinushey Seurat responded (see issue here) that they are unable to change their DESCRIPTION file. Can rsconnect update code to account for this situation? Thank you
Do I need to submit an issue to the packrat repo?
Yes, I think this would require a patch on the packrat side.
I have a similar issue with package colorBlindness
, which I installed from CRAN.
Hi all, Thanks a lot for developing this package to easily deploy Shiny apps. I have had no problems in the past when deploying apps to shinyapps.io but it now seems that I'm encountering an issue related to this.
I am also using Seurat and deploying it within the app but what seems to be crashing in my case is the package sctransform
.
When deploying I get the following error message
Error: Unhandled Exception: Child Task 807554307 failed: Error parsing manifest: Unable to determine package source for Bioconductor package sctransform: Repository must be specified
From what I have been able to gather this is due to sctransform
depending on the bioconductor package multtest
. Therefore, in the package DESCRIPTION they need to add biocViews
as is the case with older Seurat
versions. More recenet Seurat versions don't have this description and thus don't give a problem.
I am declaring options so that bioconductor repositories are within the scope of the environment via:
options(repos = BiocManager::repositories())
When trying to identify what might be going wrong I run:
rsconnect::appDependencies()
Where sctransform appears as such:
126 sctransform 0.3.1 Bioconductor
My guess is that it is trying to install sctransform
from Bioconductor when it is actually a CRAN package.
Unfortunately, I cannot roll back and use a previous version of Seurat in this case as suggested in previous comments in this issue. Is there any way to specify that it should be installing sctransform from CRAN instead?
Thank you for your time, Marc
I believe if you delete the biocViews line from the installed packages description file it will resolve the issue.
Thanks @slopp , Deleting the line solved the issue I was getting.
I also posted an issue in the sctransform
repository here.
There Christoph says the biocViews
is not needed in the package and therefore it can be removed.
Of note - for other people with this issue one needs to remove the biocViews:
line from the DESCRIPTION package of the file. To do so you can run: installed.packages()
to get the path where sctransform
is installed.
You can then head to the specified path, access the sctransform directory and modify the DESCRIPTION file there.
I met similar problems, I used install_github("the problematic package") to reinstall the package and succeeded.
Judging from https://github.com/rstudio/packrat/issues/591, this is now fixed in the dev version of Seurat.
Hello. I am getting the following strange error, apparently the Bioconductor repo error during the parsing manifest step, but the package indicated is not a Bioconductor package.
The strange thing is, Seurat (at least the modern / current version that I am using) is not a Bioconductor package. Older versions of Seurat, I believe, were on Bioconductor. Could it be caused by Seurat's dependencies or something? This worked on previous versions of rsconnect, such as this version:
devtools::install_github("rstudio/rsconnect", ref='737cd48', force = T)
which is now deprecated... It also works locally no problem.I ran the command
options(repos = BiocManager::repositories())
which works for other errors caused by true Bioconductor packages, but the error above persists after doing so.Here is my sessionInfo():
Please let me know if there is any other info I should provide. Thank you very much.