Thanks for the really useful video! Really instructional and helpful. For those that can't seem to get the plot to run, there is a step not captured in the video where the alzheimer's status is converted from string to a factor/numerical value. If this isn't done, generating the colour column returns an error. I used "pcs$Alzheimers
I have two suggestions. 1. you can use --make-bed flag, so your .ped file is not that large and it will take shorter time to run. 2. you can use "sort -gr" to sort "unadjusted" file's p-value, but of course, the result in "adjusted" file is what we want.
Hi, I tried the same data in PLINK using R studio instead of GPLINK. I am however not knowing how to do the PCA and obtain the plots; as well as the eigen values. Please help with the steps
and when I type plink in terminal of my kali linus, after showing version etc box, it says "segmentation fault"... Can you please help me with this.. Thankyou
If I'm not wrong, the sixth column in your *.ped file stores the target phenotype for association testing, no matter if it is categorical (case/control) or numerical. Maybe you could generate a *_quantitative.ped file, putting the quantitative phenotype in the mentioned column. Though I'm pretty sure there is a less cumbersome way to do this, I think it may work... As she said, you can perform association tests with phenotypes found in the *.covar file. In the *.covar file provided in the description, there both categorical and quantitative traits are included. You could try with the material available.
@@inigotellaetxeelorriaga5292 Thank you so much for the comment. You are correct. PLINKs assoc command (and other supported methods for quantitative traits) automatically realises it's a continuous variable, and perform the relevant tests. It's that easy! Also we can provide a phenotype file with many phenotypes in columns and perform tests for all of them at once with --pheno-all tag.
Really nice video! Thank you! I just have a question, in case sb else has encountered. I am running Quality Control and Association analysis per chromosome. And some chromosomes have a lambda higher than 1, whereas others don't, hence should I exclude any Principal components from the analysis? Thanks in advance!
Thanks a lot. It's really helpful. I was wondering whether you can share the name of package you used in r and also the test data to get inflation value.
I updated my system to macos catalina 10.15.4, and it seems like my plink1.07 cannot be executed locally, I googled the problem and i thought maybe it's the new system dosen't support the old plink1.07? Does anyone meet the same problem as me?
Thanks for the insightful video. When I try to plot the PCA in R, I got the following message after entering pcs$colour = palette[as.integer(pcs[,1])] Warning message: NAs introduced by coercion I entered the R script exactly as shown in your video. Any idea what might cause this problem?
This means that it cannot convert the first column in the dataframe pcs to integers. I encountered this too. I replaced the first column with integers based on if it's a case or control. Hope this helps!
Basically, analyzing only autosomal chromosomes is easier. There are methods to include the sex chromosome in GWAS but it will complicate the analysis methods you should use because of the unique nature of this chromosome.
Hi, thank you for this video. I'm going to run a GWAS for my PhD soon so wanted to run through with the plink test data set so that I can understand the pipeline. I've run into an issue with the pca step and wondered whether you might be able to assist me please? This is the error message I get: ** Unused command line option: --pca ** Unused command line option: var-wts ERROR: Problem parsing the command line arguments. This is the command script that I've used: plink --file C:\Users\Charlotte\PLINK\plink_test_gwas\crytic --pca var-wts --recode --out C:\Users\Charlotte\PhD\PLINK\plink_test_gwas\pca_2 --gplink
Thank you @math et al. Can you please make a video on Plink describing how to perform different things from Command prompt (Windows). How to write flags together? Something starting from more basic to concluding GWAS analysis. Thanks alot in advance Regards Dr Aman, PhD
It's pretty easy on Windows. Just download from plink 1.9's website the windows version, and open CMD in that directory where plink.exe is. Then you can use plink from that directory if you don't want to add it to PATH. And for how to write flags, it works just like any other command line program. The docs for plink are quite good too! Hope this helps
Thank you! The video is very helpfull! (: I'm having trouble executing plink in my mac (using macOs Big Sur 11.2.3). I'm using the 1.9 version for 64 bit mac. But plink shell keeps running and finishes with a "[Process completed]" and the shell does not accept any command. Anyone having the same problem as me? Any help? :(
Thanks for the really useful video! Really instructional and helpful.
For those that can't seem to get the plot to run, there is a step not captured in the video where the alzheimer's status is converted from string to a factor/numerical value. If this isn't done, generating the colour column returns an error. I used "pcs$Alzheimers
thank you!!!! would absolutely love more vids on PLINK, R etc.
thanks! very helpful for beginners.
holy Cow! :) this is some cool sh**. thank you for sharing this knowledge with the world.
I have two suggestions.
1. you can use --make-bed flag, so your .ped file is not that large and it will take shorter time to run.
2. you can use "sort -gr" to sort "unadjusted" file's p-value, but of course, the result in "adjusted" file is what we want.
Hi, I tried the same data in PLINK using R studio instead of GPLINK. I am however not knowing how to do the PCA and obtain the plots; as well as the eigen values. Please help with the steps
Thanks for the insight
Can you please show your plink code?
I am not able to run gPlink on my macos please help. it says cannot run commands locally
Lambda GC between 0.95-1.05 is benign. You included the PC1 and seems to me is overcorrected population structure
and when I type plink in terminal of my kali linus, after showing version etc box, it says "segmentation fault"... Can you please help me with this..
Thankyou
can i use GWAS for viral studies? Like if i wanted to study SNPs in various covid samples, could i use this approach?
and when I click on gPlink.jar file, it opens extractor and ask where to extract 4 number of folders in it..:(
I downloaded the plink 107 then opened it, but my ped and map files wasn’t recognised by plink.
Great video!
This is awesome!
What should be changed in this workflow when you have a quantitative phenotype instead of case/controls?
If I'm not wrong, the sixth column in your *.ped file stores the target phenotype for association testing, no matter if it is categorical (case/control) or numerical. Maybe you could generate a *_quantitative.ped file, putting the quantitative phenotype in the mentioned column. Though I'm pretty sure there is a less cumbersome way to do this, I think it may work... As she said, you can perform association tests with phenotypes found in the *.covar file. In the *.covar file provided in the description, there both categorical and quantitative traits are included. You could try with the material available.
@@inigotellaetxeelorriaga5292 Thank you so much for the comment. You are correct. PLINKs assoc command (and other supported methods for quantitative traits) automatically realises it's a continuous variable, and perform the relevant tests. It's that easy! Also we can provide a phenotype file with many phenotypes in columns and perform tests for all of them at once with --pheno-all tag.
@@generic395 I didn't know about the --pheno-all tag, thank you!
What is the input data for PLINK that will be processed
Really nice video! Thank you! I just have a question, in case sb else has encountered. I am running Quality Control and Association analysis per chromosome. And some chromosomes have a lambda higher than 1, whereas others don't, hence should I exclude any Principal components from the analysis? Thanks in advance!
Since you are doing everything on 1 chromosome at a time, couldn't you do the PCA for each chromosome to determine if you need PCs as covariates?
Thanks a lot. It's really helpful. I was wondering whether you can share the name of package you used in r and also the test data to get inflation value.
can u explain how should we take values for maf mind and geno
I updated my system to macos catalina 10.15.4, and it seems like my plink1.07 cannot be executed locally, I googled the problem and i thought maybe it's the new system dosen't support the old plink1.07? Does anyone meet the same problem as me?
I know I've used plink 1.9 on Catalina so you can use that instead. Make sure that the program is 64 bit and not 32, that might be your problem!
Thanks for the insightful video. When I try to plot the PCA in R, I got the following message after entering
pcs$colour = palette[as.integer(pcs[,1])]
Warning message:
NAs introduced by coercion
I entered the R script exactly as shown in your video. Any idea what might cause this problem?
This means that it cannot convert the first column in the dataframe pcs to integers. I encountered this too. I replaced the first column with integers based on if it's a case or control. Hope this helps!
Brother, there are multiple files like eigenvec,eigenvaland etc., which one needs to be changed and how?
Thank you
@llub888
Why do you exclude the sex chromosomes in your analysis?
Basically, analyzing only autosomal chromosomes is easier. There are methods to include the sex chromosome in GWAS but it will complicate the analysis methods you should use because of the unique nature of this chromosome.
how can i filter the files
Hi, thank you for this video. I'm going to run a GWAS for my PhD soon so wanted to run through with the plink test data set so that I can understand the pipeline. I've run into an issue with the pca step and wondered whether you might be able to assist me please?
This is the error message I get:
** Unused command line option: --pca
** Unused command line option: var-wts
ERROR: Problem parsing the command line arguments.
This is the command script that I've used:
plink --file C:\Users\Charlotte\PLINK\plink_test_gwas\crytic --pca var-wts --recode --out C:\Users\Charlotte\PhD\PLINK\plink_test_gwas\pca_2 --gplink
Thank you @math et al. Can you please make a video on Plink describing how to perform different things from Command prompt (Windows). How to write flags together? Something starting from more basic to concluding GWAS analysis. Thanks alot in advance
Regards
Dr Aman, PhD
It's pretty easy on Windows. Just download from plink 1.9's website the windows version, and open CMD in that directory where plink.exe is. Then you can use plink from that directory if you don't want to add it to PATH. And for how to write flags, it works just like any other command line program. The docs for plink are quite good too! Hope this helps
really useful
Thank you! The video is very helpfull! (:
I'm having trouble executing plink in my mac (using macOs Big Sur 11.2.3).
I'm using the 1.9 version for 64 bit mac. But plink shell keeps running and finishes with a "[Process completed]" and the shell does not accept any command. Anyone having the same problem as me? Any help? :(
Thank you Daniela!! (I assume its your name)
Um i was looking for the music game but okay