[AsoBioInfo]

The check for the rating is already included. The marker is only one (because the pair/column is one)may be, that's the reason for not showing LD - just a guess

[gringer]

Yes, of course. For some reason I'd forgotten that you had only one marker -- no other markers to compare with, so no LD checks are possible (LD is a pairwise measure).

[AsoBioInfo]

Thanks!

Hmmm... So the haplotype possibilities will be (0,0),(0,1),(1,0),(1,1). Right?

[gringer]

Hmmm... So the haplotype possibilities will be (0,0),(0,1),(1,0),(1,1). Right?

Those are the genotype possibilities for a single polymorphism at a diploid locus, with alleles coded as 1/0.

Upon further reflection, my previous answer of four was incorrect (which I've now changed to reduce confusion from other readers)... there are only two possible haplotypes (ways of describing the sequence on a single chromosome) at such a locus: 1 and 0. You need multiple markers to get non-trivial haplotypes.

[emilyjia2000]

Hi gringer,

I tried to use Haploview as well, in Linkage Fromat, I loaded PED file to the Data File, and MAP file to Locus Information File, it gave the error message:

"File Error: Individual 6 in family L3 appears more than once".

I have 4 sisters all in the same family L3, the first 6 columns are same, I don't if this error happened due to that?

Thanks for the help.

[gringer]

If you have pedigree information, then that should be in the first four columns. Individual IDs within a family need to be unique. If you've got four sisters, then it should look something like this:

L3  1  0 0  1  0  A A  G G  A C 
L3  2  0 0  2  0  A A  A G  0 0 
L3  3  1 2  2  0  A C  G G  A C 
L3  4  1 2  2  0  A A  G G  A A 
L3  5  1 2  2  0  A A  G G  A C 
L3  6  1 2  2  0  A A  A G  C C

Notice that the second column is unique within this family (different for each line). That is what Haploview is complaining about.

[emilyjia2000]

Thanks gringer for your quick response. the family size is quite big, there are 16 members in the family, I personally separated the big family into three small ones, I am not sure if it is right? one small branch family has 3 generations, the second one has 2 generations, the third one has 1 generation (as I said, only 4 sisters).
For the pedigree ID, I gave L1, L2, L3, in each family, set up individual ID with unique one, for example
L1 10 0 0 1 1
L1 9 0 0 2 1
L1 2 10 9 1 2
L1 3 10 9 1 1
L1 4 10 9 1 1
L1 5 2 0 2 2
L1 6 2 0 2 2
L1 7 2 0 2 1
L1 8 2 0 2 1
L2 1 0 0 1 2
L2 15 1 0 1 1
L2 16 1 0 1 1
L3 11 0 0 2 1
L3 12 0 0 2 1
L3 13 0 0 2 1
L3 14 0 0 2 1

Or construct the pedigree as a big family, like
L1 10 0 0 1 1
L1 9 0 0 2 1
L1 2 10 9 1 2
L1 3 10 9 1 1
L1 4 10 9 1 1
L1 5 2 0 2 2
L1 6 2 0 2 2
L1 7 2 0 2 1
L1 8 2 0 2 1
L1 1 0 0 1 2
L1 15 1 0 1 1
L1 16 1 0 1 1
L1 11 0 0 2 1
L1 12 0 0 2 1
L1 13 0 0 2 1
L1 14 0 0 2 1

I used the second one to try it on Haploview, it has error message again with "Files contain zero valid individual."

Please see the attached pedigree structure.

Thanks a lot for the help.

[gringer]

The error you mentioned was the following:

"File Error: Individual 6 in family L3 appears more than once".

This error suggests a duplicate individual in family L3. Perhaps you've defined that family twice.

If you have access to a Linux command line, could you try the following command?

grep '^L3[[:space:]]\+6' <PED file>

[emilyjia2000]

Hi gringer,

in pedigree structure, the parents info must have both father and mother? for some subfamily, I only have either father or mother, not both of them. In that case, missing one of the parents could cause error?
Thanks a lot

[gringer]

in pedigree structure, the parents info must have both father and mother? for some subfamily, I only have either father or mother, not both of them. In that case, missing one of the parents could cause error?
Thanks a lot

It's possible, but that shouldn't produce an error saying no lines have any correct information, or an error saying there is a duplicate line. I suspect that there's some other non-obvious error in your input file that is causing problems, but can't break it down more than that if you have no way to test things on a command line.

[emilyjia2000]

You are right, I found the fatal error by running merlin: "Mendelian inheritance errors detected". It looks like ped file generated by vcftools from VCF file is not correct. Do you have any idea how to correct it?

Thanks gringer for the help.

[gringer]

Do you have any idea how to correct it?

No, sorry. I would be tempted to get rid of the pedigree information from the PED file (i.e. set everyone's mother and father to 0).

I think a Mendelian inheritance error means that it's incredibly unlikely that the observed parental genotypes could cause the observed child genotypes. One example would be a child with 'A A', and both parents 'C C'. I suppose if you've got enough columns in your linkage file, the probability of such errors for every individual approaches 1.

These should really be warnings, rather than errors, because when you're playing around with thousands of genotypes, rare events are quite frequent. But then again, Merlin is not all that great when working with thousands of loci.

[emilyjia2000]

I probably figured out the problem, when I converted VCF to PED by vcftools, I didn't use one chromosome, which could mess up it. I am not so sure, but it could be one reason. The problem is that when I tried to use vcftools convert VCF to PED with only one chromosome, say chr4, nothing's generated, I don't know what's wrong?

Thanks for the help

[alizaid12]

Hello. I chanced across this forum and I wonder if I can ask a rather naive question. I recently used PLINK/ Haploview for understanding the data of the highly ranked SNPs in a GWAS analysis. I had planned to visualize my SNPs in LD blocks [and their proximity to genes of interest[, by using Haploview. Additionally, I used dbSNP to confirm the proximity of my SNPs to genes of potential interests.

Curiously enough, no matter how hard I have tried, I cannot get Haploview to show the same genes, that I can visualize when using dbSNP. I have tried repeatedly but with no satisfactory breakthrough. I would be so grateful if someone could have some suggestions on what I may be doing incorrectly.

I can also give a list of sequential steps [ as to what I am doing that is giving different results]

Thankyou for taking the time to read my question.

A.Z

[gringer]

Your questions are a bit unclear. Could you try being a bit more specific?

First, How do you want your data to look? How many genes / chromosomes do your data cover? I've attached a few examples from my thesis. For the ones with the D' squares, I've used the SVG export function of Haploview and manually added in the genes and tweaked the image using Inkscape. The chromosome plot is an alternative plot that shows the whole genome at a glance, based on figures in the first WTCCC GWAS paper.

Second, what have you tried doing? If using the information track download of Haploview, have you checked to make sure that the build number is correct?