Error in matrixStats::colMedians(x, rows = rows, cols = cols, na.rm = na.rm, Argument 'useNames' must be either TRUE or FALSE

[zeinabrammal]

I am performing DNA methylation analysis using 'minfi' and 'dmapaq' in R studio but i am consently having this error. How can i fix it? I tired to downgrade the matrixStats versin to 1.1.0 but the error remains.

Error in matrixStats::colMedians(x, rows = rows, cols = cols, na.rm = na.rm,  
Argument 'useNames' must be either TRUE or FALSE  

 packageVersion("matrixStats")
[1] ‘1.5.0’

> packageVersion("minfi")
[1] ‘1.44.0

I appreciate your help

[yaccos]

Do you use the latest version of DelayedMatrixStats (3.20)? The most recent version of it should work flawlessly with matrixStats 1.5.0, wheras using a new version of matrixStats and an old version of DelayedMatrixStats at the same time is likely to cause problems.

[yaccos]

If upgrading DelayedMatrixStats does not work, please provide the output of traceback() after the error occurs.

[zeinabrammal]

i'm trying to update DelayedMatrixStats becuse im working on the 1.20 version, but it's not working because of an old version of Bioconductor, and I'm working on a shared server, once I solve this I'll make an update.
Thanks!

[NWU-psp]

If upgrading does not work, please provide the output of after the error occurs.DelayedMatrixStats``traceback()

Hi, I've been having the same problem recently. Errors are reported both when using quasi-timing analysis with monocle3 and when using slingshot，My error is as follows:

> cds <- align_cds(cds, alignment_group = "Sample")
Aligning cells from different batches using Batchelor.
Please remember to cite:
	 Haghverdi L, Lun ATL, Morgan MD, Marioni JC (2018). 'Batch effects in single-cell RNA-sequencing data are corrected by matching mutual nearest neighbors.' Nat. Biotechnol., 36(5), 421-427. doi: 10.1038/nbt.4091
Error in matrixStats::colVars(x, rows = rows, cols = cols, na.rm = na.rm,  : 
  Argument 'useNames' must be either TRUE or FALSE

In addition, when using slingshot, the error is as follows, although he did not specify which package reported the error, but I think it must be matrixStats package：

> sim <- slingshot(sim, 
+                  clusterLabels = 'celltype4', 
+                  reducedDim = 'tsne'
+ )
Error in if (!useNames) { : missing value where TRUE/FALSE needed

This is my version of the package：

> packageVersion("matrixStats")
[1] ‘1.5.0’
> packageVersion("DelayedMatrixStats")
[1] ‘1.22.6’

DelayedMatrixStats package does not seem to be used in this algorithm, so I do not understand whether it needs to be updated?

I would appreciate your prompt reply to me ！！！！ This has been bothering me for a few days because I feel like I should have accidentally updated a few other R packages the last time I installed them (maybe including the MatrixStats package?). ...

[yaccos]

monocle3 does indeed utilize DelayedMatrixStats, so it leaves you with the option of either downgrading matrixStats (to version 1.1.0) or upgrading DelayedMatrixStats, I recommend the latter.

[zeinabrammal]

THIS IS AN UPDATE!
I used an older version on matrixStats because apparently the problem was a packages incompatibility and it worked.
I used this command to install the older version:
remotes::install_version("matrixStats", version = "1.1.0")

[yaccos]

Sure enough, downgrading matrixStats is an option if the Bioconductor packages cannot be upgraded. Reading bug reports, it appears that matrixStats has a hidden package compatibility problem. matrixStats by itself support older versions of R back to 3.4.0 (prior to matrixStats version 1.5.0 it even supported versions of R back to 2.12.0). However, each new Bioconductor release appears to target the most recent version of R, meaning that you cannot upgrade Bioconductor packages on old versions of R. CRAN packages do not have this limitation, hence users risk to break Bioconductor packages such as DelayedMatrixStats when running on an old version of R and installing the most recent version of matrixStats

[NWU-psp]

OK！ Thank you！