Unable to install new version of DropletUtils

[liron27]

Hi,
I have tried to install the new version of the package using:
devtools::install_github("MarioniLab/DropletUtils")

and got the following error:

ERROR: compilation failed for package ‘DropletUtils’

• removing ‘/tmp/Rtmp4VjwFR/Rinst19fd655306c7/DropletUtils’

Thanks

[LTLA]

Try installing using the recommended instructions on the Bioconductor landing page.

[liron27]

I did but it doesn't include the new function removeAmbience

…

On Tue, Sep 29, 2020 at 12:42 PM Aaron Lun ***@***.***> wrote: Try installing using the recommended instructions on the Bioconductor landing page. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKS5HHAHBZ2AO3VAJBMTSIIFAFANCNFSM4R6A3SDA> .

[LTLA]

You can either wait a few hours for the new builds to propagate on Bioconductor. Or you can give me much more detail about the entire error log, I can't figure out what's going on from just the last line.

[liron27]

I just copied the removeAmbience function from github to my local directory and then I used the following code:

raw.count <- Read10X(data.dir = "path.to.raw.count")
ambient <- estimateAmbience(raw.count)
x<-removeAmbience(raw.count,ambient, colnames(raw.count))

I am not sure if this is the way to go, but it is still running (for almost 20 minutes now...)

[LTLA]

Stop, stop stop. Hold up. You really need to give it clusters. Did you read the documentation?

Install the package properly rather than doing whatever you're doing.

[liron27]

I read, but have few issues: 1. Doing the clustering on the original raw data (without cleaning ambient rna) is computationally heavy - my original raw data has over 6 millions barcodes ... 2. Should I run the entire pipeline of normalization, PCA, clustering on the raw data and then bring it back to “removeAmbience”? Thanks!

…

On Tue, Sep 29, 2020 at 3:41 PM Aaron Lun ***@***.***> wrote: Stop, stop stop. Hold up. You really need to give it clusters. Did you read the documentation? Install the package properly rather than doing whatever you're doing. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKSYMEVCA5HDJ5KIBNM3SIIZ6VANCNFSM4R6A3SDA> .

[LTLA]

I doubt you have 6 million cells generated from a single run of a 10X machine. The maximum number of barcodes is around 700k, at least for versions 2 and 3 of the 10X chemistry.

[liron27]

Yep, I suspected it’s an issue... I am using the “Read10X” function with a path to the directory that contains the barcodes, features and matrix.

…

On Tue, Sep 29, 2020 at 3:59 PM Aaron Lun ***@***.***> wrote: I doubt you have 6 million *cells* generated from a single run of a 10X machine. The maximum number of barcodes is around 700k, at least for versions 2 and 3 of the 10X chemistry. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKSYJZ3OATAXTG6HTW3LSII4D7ANCNFSM4R6A3SDA> .

[LTLA]

The book build crashed so you won't be able to see the latest chapter of the book, but you can just read it from the Rmarkdown starting at:

https://github.com/Bioconductor/OrchestratingSingleCellAnalysis/blob/092880538f4c475fd397c6ab390a69027c545cdc/droplet-processing.Rmd#L137

Or you can just wait for Friday, hopefully the book will have built by then.

[liron27]

Thank you!

…

On Tue, Sep 29, 2020 at 8:54 PM Aaron Lun ***@***.***> wrote: The book build crashed so you won't be able to see the latest chapter of the book, but you can just read it from the Rmarkdown starting at: https://github.com/Bioconductor/OrchestratingSingleCellAnalysis/blob/092880538f4c475fd397c6ab390a69027c545cdc/droplet-processing.Rmd#L137 Or you can just wait for Friday, hopefully the book will have built by then. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKS423ASKTP3PXJY6FVLSIJ6TVANCNFSM4R6A3SDA> .

[liron27]

Regarding the 6 million barcode issues from 10x - 10x increased the barcode # in the whitelist in V3 version to roughly ~3.5 Million BCs. I believe the barcodes.tsv.gz contains barcode sequences and their reverse complement sequences, so making the total # ~6.8M. On Tue, Sep 29, 2020 at 11:19 PM Liron Grossmann <[email protected]> wrote:

…

Thank you! On Tue, Sep 29, 2020 at 8:54 PM Aaron Lun ***@***.***> wrote: > The book build crashed so you won't be able to see the latest chapter of > the book, but you can just read it from the Rmarkdown starting at: > > > https://github.com/Bioconductor/OrchestratingSingleCellAnalysis/blob/092880538f4c475fd397c6ab390a69027c545cdc/droplet-processing.Rmd#L137 > > Or you can just wait for Friday, hopefully the book will have built by > then. > > — > You are receiving this because you authored the thread. > Reply to this email directly, view it on GitHub > <#52 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AQRAKS423ASKTP3PXJY6FVLSIJ6TVANCNFSM4R6A3SDA> > . >

[LTLA]

I would be shocked if all 6 million barcodes had counts. Even looking at the existing output that I get from CellRanger 3, I only get non-zero counts for ~600k barcodes.

[liron27]

This is from 10x genomics I’m using the makecount command to generate both the unfiltered and filtered matrix from the fastq file.

…

On Thu, Oct 1, 2020 at 12:51 PM Aaron Lun ***@***.***> wrote: I would be shocked if all 6 million barcodes had counts. Even looking at the existing output that I get from CellRanger 3, I only get non-zero counts for ~600k barcodes. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKS6XRAPZ5NABWRF3BRLSISXSHANCNFSM4R6A3SDA> .

[LTLA]

Did you resolve your issue?

[liron27]

No

…

On Sun, Oct 11, 2020 at 9:34 PM Aaron Lun ***@***.***> wrote: Did you resolve your issue? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRAKS2XMWYUQS2TLEU5R3DSKJMJ3ANCNFSM4R6A3SDA> .