ฝัง
- เผยแพร่เมื่อ 31 ธ.ค. 2021
- I cover the basics of installing and using Cell Ranger on a 10x single-cell RNAseeq data. I show basic usage and briefly cover run QC. Cell Ranger creates the count table and other outputs which will be used in downstream single-cell analysis.
![เกิดใหม่ทั้งทีก็เป็นสไลม์ไปซะแล้ว ซีซั่น 3 - ตอนที่ 57 [ซับไทย]](http://i.ytimg.com/vi/pNhft3YZFRE/mqdefault.jpg)
Just want to show my appreciation to you for making this tutorial. A big life saver, thank you thank you thank you thank you thank you
Thank you for this amazing tutorial!!
You talked about the expected cells and how this info is important, why it's important? And how researchers decide what number they wanna go for?
Thank you again!!!
Hi. You pick that number when you are designing your experiment. Usually it is in the range of 2-10k. It is important because you are telling the program roughly how many cells you expect, so it can better call what are real cells versus artifacts. I think more recent versions of cell ranger do an even better job finding the threshold without an accurate expected cell input, but it is still important to have a basic understanding of how many cells you expect to have in case something goes wrong with the experiment.
Hey, good abstract about the cell ranger pipeline. I have 2 questions. First, Is not longer available the Barcode Rank Plot in the cell-ranger2.0 html report? because, I have not available this report in my html report ... I had to built my own plots. The other question is, Is the configuration to create the bc filtered matrix defined automatically ? ... I mean, How are defined the cutoffs to create this bc filtered matrix? .... Cheers
Thank you so much! Would it please be possible to do a video for Cell Plex? To use Cell ranger multi function please! Also can you please tell me which computer are you using because running cell ranger hardware requirement is quite stringent.
Hi. I have yet to use Cell Plex, sorry. I am running it on 24cpu, 128gb ubuntu system. As far as I am aware the only requirements for cell ranger are ~>32 gb of ram and 1 cpu. Not sure all the operating systems it CAN work on, but I have never had issues with ubuntu or centos.
Hi, thank you for the useful video! One question: I have 1 sample, several lanes per sample as well as several flow cells for this sample. Do I have to merge/cat some files?
You went for a lot of depth on this sample, eh? I've never had this issue before so i'm not 100% sure. If they are from the same library then I think you can probably get cell ranger to run on all the files. I don't think you have to concat them. You may have to modify the file names a little though to trick it
how are the antisense mapped read dealt with in the count matrix> is it discarded or token into account as an extra read within the gene, ( if its something that u adjust and choose , what do u recommend to do)
It is a nice tutorial. Do you know (or have any reference) how well recommended is to use cell ranger count vs SALMON or any other mapper plus counter software?
Not off hand. I would say if it is 10x data - you should probably just use the 10x cellranger pipelines. If there are alternatives, I don't see them making a big difference in my opinion.
I think that it is not efficient to use cell ranger data with pseudo-aligners. But you can use faster ones like STARSolo and Minimp2.
Hey, awesome video. I am an undergrad who helps in a lab at my school, and I have a 10X dataset that my PI gave me to learn this process with. I only have fastq files. The cells are Xenopus tropicalis embryonic cells. How would I create a reference transcriptome for this data? Also, the data I was given lacks files that have "I1" or "I2" in their names. I only see "R1" and R2". What are these files that I am missing? Are they necessary for using cellranger? Your videos are extremely helpful--us undergrads w/o much experience appreciate them greatly btw.
You'll need to build the reference with 10x mkref using this: useast.ensembl.org/Xenopus_tropicalis/Info/Index
And don't worry about not having the I1 or I2 files
Hi Mark. Thank you for this video! May I know where to get the fastq files in this tutorial?
Hi Chris, these were fastq from one of my collaborators. Don't think they are published yet, sorry!
@@sanbomics Thank you for your quick reply! Just want to follow along your tutorial, I can use any fastq file, is that correct?
Any fastq files that were processed by 10x. If you get them online you will likely have to rename them using the 10x standard naming format.
This video has been so helpful! I'm curious if you could do a follow-up on using the Cell Ranger `aggr` function following `count` - when to use, what the outputs are, and how to compile them downstream. (Or comment here :)
I'm confused because I thought `aggr` would compile all the samples I have from individual `count` runs into a single matrix (?), but it seems I still have individual outputs for each sample. Is compilation just done in Seurat?
I'm not really a fan of aggr for RNA especially. You can just combine/integrate the samples later. and much quicker
1.Is this just one sample?What does the name of V_I_48_3_S2_L001_I1_001.fastq, V_I_48_3_S2_L001_I2_001.fastq mean? (I1-index, R1-cell barcode seq, R2-cDNA seq) Does the L001 from one illumina machine, and L002 from another illumina machine? Those are one single sample of S2?
2. How can one sample's analysis finally end in a fold change of a specific marker? Does the analysis compared to the reference basic level?
3. From my understanding, we need to get the UMI out of each sample, and do a expression analysis (e.g., with DESeq2) with at least two level of samples (e.g., Con vs. Trt)?
4. Thank you very much for your answer, I just started to learn do deal with those data.
1) Just one sample. Depending on your instrument samples can have many fastq files. L001 denotes lane. For example a 4 lane novaseq run can have 8 fastq per one sample: L001, L002, L003, and L004. You don't need the Index files (I1, I2) for analysis but they are often together with the fastq. V_I_48_3 is the sample name; S2 is the sample number; L001 is the lane; I1 is the forward read index; R1/R2 would be forward/reverse fastq; 001 is the file number--very rarely is this not 001
2) Cellranger produces a count matrix of UMI for every cell and gene. You will need to process that further after you run cellranger. If you like R use seurat. If you like python use scanpy. I have a scanpy tutorial.
3) You want to do DE between two single-cell samples? In that case it is probably best to do pseudo-bulk. But you can also also use the build in seurat or scanpy functions as well on an integrated dataset. If you give me more details about what you are trying to do I can give you more advice.
4) We all started somewhere! Glad I could help. Please let me know if you have any more questions.
Thank you very much for all your answers. I will take your advice especially in question.
Hi, how did you specify in your pipeline that you wanted to include the analysis?
It is included by default. You can turn it off but it really doesn't add much time at all so I like to keep in in there just for QC. The output file is the html file, which is in the outs folder I think
That makes sense, thank you so much for the response!
Thank you so much for the amazing tutorial! i use spaceranger count and it runs but my ubuntu stops due to windows not having enough disk space. I tried deleting via ubuntu and windows but the windows space doesn't increase. Can I get some help?
Sounds like you are using a ubuntu VM in windows? It sounds like you need to recreate the image with higher disk space
Thank you so much! But I am trying to run the cellranger on Mac OS? but failed, zsh: exec format error. Can I get some help?
Hmm. I'm not sure if it will work on Mac. Best to find access to a linux machine if possible. Good luck!
Sir i am getting this error while executing count. please help.
Permission denied: couldn't open MRO file
hmm never come across that before. Try running it as sudo
How do you fix the issue "curl:(56) Failure when receiving data from peer"
Hi! that's an interesting error. You copied and pasted the link exactly as they have it in on their downloads page and it wasn't clipped or anything? Maybe it was transient, have you since had success or is it still giving you issues?
@@sanbomics I am able to use the first link however the second always results in the curl: (56) error. I am trying to use the mouse reference specifically. I tried the human just to see and that one resulted in the same error as well. Any advice?
You can try wget instead. Or try it on another computer/network altogether. If after that it is still not working, maybe we can work something out where I share the file on google drive or something.
Hi, sir can you please tell from where we can easily download sc rna datasets except ncbi geodatasets
The easiest way is to find a paper you are interested in and check the data availability section. But other than geo, EMBL-EBI is very easy to use.
@@sanbomics thank you.
super
could u share the code below.
Hi, you just want the command I used to run cellranger for this video? Here it is, but make sure to replace everything accordingly!
cellranger-6.1.2/cellranger count --id tutorial_sample --transcriptome refdata-gex-GRCh38-2020-A/ --fastqs Sample_T_I_48_3/ --sample T_I_48_3 --expect-cells 10000 --localcores 20 --localmem 100