What would cause you to have no phenotype data?
I have a VCF file that I am running through PLINK to get PED and MAP files, and then .assoc files, so I can get Manhattan plots. Whenever I try to get any files from PLINK it says there are no phenotypes present. It is a plant genome.
What could be causing this? I started with Fastq files, ran BWA on them to align them with the reference, and then marked duplicates in sandblaster, converted sam to bam, and then sorted and indexed it, and ran it through freebayes to get the VCF.
everything is the right size file, and ran normally. What might be the problem?
Why would someone have no phenotypes?
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
GWAS from direct samples... is imputation necessary?
written by sabrilo171 •
-
genome data size
written by kkk •
-
GWAS and vcf files. Lacking Phenotype
written by BQ •
-
plink case/control association analysis and Manhattan plot
written by nayeona.hi
-
Burden Test for whole exome sequencing data running into errors
written by shraimrawan •
-
How to subtract out wildtype variants, and what is it?
written by tmrhyd •
-
Unable to find FASTA index entry for 'unitig_2993_pilon error.
written by tmrhyd •
-
PLINK assoc input: can it be two vcf files case/control?
written by Pin.Bioinf
-
GWAS result with pval equal zero. How to handle and how to plot?
written by GabrielMontenegro
-
Trouble reading plink (.raw) file into R (adegent)
written by sryan6 •
Well, what are your phenotypes?, i.e., what are you hoping to compare? Is it plant species A versus plant species B? Why would you try to generate association test (.assoc) files with PLINK on just a VCF without specifying what are your sample groups?
Perhaps you should explain better what you are hoping to achieve. Based on the minimal information that you have provided, there should be no need to perform an association test in PLINK. Please show also the PLINK commands that you have been using. Also, please read the PLINK manual pages so that you better understand how to use it.
Essentially: what is the biological question you are trying to answer?
All I know are they are different mutations of the same algae, and I am hoping to compare the snps of them, so I need to get Manhattan and q-q plots. It is Algae A wild type vs Algae A mutation1 ect, mapped against the chloroplast as a reference. How would I specify my sample groups in a VCF?
the plink commands I have been using are :
the above was to create the MAP and PED files
this one was to create the .assoc file. But I get the "no .assoc made because there are 0 phenotypes error each time.
I have been reading over the PLINK manual, and haven't found anything to help this issue. is there a specific part of the manual that I should be reading?
Thank you so much for your help!
If it is a 1 vs 1 comparison, then you cannot make statistical inferences on that. If that is the case, you just need to do a very simple comparison between the samples in your VCF, for example, checking if a variant is present in either A or B, or both.
With the code that you have written, PLINK will never know what are your sample groups. You would have to create a custom FAM file and specify that in the command where you are running
--assoc. However, as mentioned, if you just have 1 sample versus 1 sample, then you cannot really use PLINK at all.I have one genomic sequence of the algae, mapped onto the chloroplast reference for that algae. As far as I know I need to get a Manhattan plot of the SNPs of the genome vs the reference chloroplast, so in essence It is a 1vs 1 comparison. I need the .assoc file from plink to make it into an Manhattan plot through. I do have 6-7 genomes I am mapping onto the chloroplast and mitochondria references (same references, different genome sequences ) do I need to get the VCF or sorted and indexed Bam files of all 7 sequences and combine them somehow to get an .assoc file to make into an Manhattan plot?
I was under the impression that you can make individual mapped genomes into Manhattan plots. Is that incorrect?
Please advise me on what I should do to get these plots with the type of data I have.
So sorry for the confusion!
Thank you
A Manhattan plot has, on the y-axis, the negative log [base 10] of a p-value; the x-axis is genomic position. You do not have any p-values and I see no clear path for you to obtain them. Are you following some tutorial or being told to do analyses by somebody else?
So, you have 6 or 7 genome sequences, and you have called variants in these against the chloroplast reference. It is still difficult to derive credible statistics from such data.
What is the ultimate aim of your study?
I am not at liberty to answer some of these questions, but theoretically, is there anything I could do to get P values?
Thank you so much and apologies for the lack of information!
Would this data be the P values? PAIRED=0.991525;PAIREDR=0;
Linked thread: Is the paired value the P value?
You have created a new question related to this current question: Is the paired value the P value?
For this reason we are closing this. This allows us to keep the site focused on the topics that the community can help with.
If you disagree please tell us why in a reply below, we'll be happy to talk about it.