search

NKI-GCF / XenofilteR

Filtering of PDX samples for mouse derived reads
GNU General Public License v3.0
27 stars 7 forks source link

Can XenofilteR be used for single cell RNA seq? #13

Closed YuliaInn closed 4 years ago

YuliaInn commented 4 years ago

Will I be able to filter out mouse reads from sc RNA-seq using your tool?

I aligned my fastq files separately to human and mouse references getting 2 bam files. My goal is to get rid of mouse cells and analyze the remaining cells.

thank you!

PeeperLab2 commented 4 years ago

Interesting question. I have never used it for scSeq but XenofilteR will distinguish between reads from mouse and human origin if you have 2 bam files.

And you could add another level by checking the number of sequence reads that are classified as human and mouse (will be 1% vs 99% or vise versa) and identify all cells that are from mouse, making the filtering even better. Or maybe simply count number of sequence reads after filtering per cell and remove all cells with very few reads.

Hope it works for you! If you have any questions feel free to send another email.

Best Oscar

On 29 Jun 2020, at 17:24, YuliaInn notifications@github.com<mailto:notifications@github.com> wrote:

Will I be able to filter out mouse reads from sc RNA-seq using your tool?

I aligned my fastq files separately to human and mouse references getting 2 bam files. My goal is to get rid of mouse cells and analyze the remaining cells.

thank you!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/PeeperLab/XenofilteR/issues/13, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AB7X5RA2JABSWQTPPGS445LRZCW3ZANCNFSM4OLKO32Q.

YuliaInn commented 4 years ago

thank you for your prompt reply. I actually ran Xenofilter on hose 2 files. But when I tried to quantify the filtered bam file in order to get the expression matrix, it gave me a matrix without cell columns. I used featureCounts from this tutorial

PeeperLab2 commented 4 years ago

Hi,

And this function does give you the matrix with cell columns when running on the original bam file? XenofilteR should leave the bam file exactly like its original minus the mouse reads. I would expect behaviour of follow-up tools to be exactly as for for the pre-XenofilteR files.

On 29 Jun 2020, at 17:45, YuliaInn notifications@github.com<mailto:notifications@github.com> wrote:

thank you for your prompt reply. I actually ran Xenofilter on hose 2 files. But when I tried to quantify the filtered bam file in order to get the expression matrix, it gave me a matrix without cell columns. I used featureCounts from this tutorial https://scrnaseq-course.cog.sanger.ac.uk/website/construction-of-expression-matrix.html#reads-quantification

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/PeeperLab/XenofilteR/issues/13#issuecomment-651203708, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AB7X5RDN4PKFTXLW3THMQVLRZCZKZANCNFSM4OLKO32Q.

YuliaInn commented 4 years ago

I guess the problem was that my data did not contain UMIs. I think my analysis should be entirely done in cellranger instead. But is there any way to see the percentage of mouse contaminated reads? I am not familiar with bam files. Is there a software to open it? I thought we could just ignore mouse cells if there are not too many of them in our sample.

PeeperLab2 commented 4 years ago

Hi,

There are many tools to work with Bam files. Samtools is most used and much can be done in Rsamtools if you would like to stay in R. Others have probably solved your issue already. I would start by looking into scSeq papers of PDX and see how they have solved the mouse cell issue. I can imagine that there are genes that are expressed specifically in mouse or human, you might be able to use this to distinguish human from mouse cells.

Good luck with the analysis! Best Oscar

On 1 Jul 2020, at 22:41, YuliaInn notifications@github.com<mailto:notifications@github.com> wrote:

I guess the problem was that my data did not contain UMIs. I think my analysis should be entirely done in cellranger instead. But is there any way to see the percentage of mouse contaminated reads? I am not familiar with bam files. Is there a software to open it? I thought we could just ignore mouse cells if there are not too many of them in our sample.

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/PeeperLab/XenofilteR/issues/13#issuecomment-652635839, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AB7X5RGCU3CEFY4ZMXW53V3RZONPNANCNFSM4OLKO32Q.

YuliaInn commented 4 years ago

thank you very much!