Interpreting Ldhat Outputs

Question

[GabrielMontenegro]

I am trying to use the LDhat package on genowide SNP phased data to estimate the population recombination rate of a region of the genome. However, after running my files using the interval and stat programes of the package, I am having difficulties in interpreting the output files. Has anyone used those programs before, and knows how to interpret them?

Here are my results based on stat (program for summarising the results from interval):

Output (on screen):

Reading data: 5000 lines (miss first 1000) of 12368 points.........Data read successfully
Mid = 2000, L95=101, U95=3900
Sample     1
Sample     2
Sample     3
Sample     4
.
.
.
Sample 12368

Average total change in rate = 272.542
Average total # changes = 657.669

res.txt (output file generated):

Loci    Mean rho    Median         L95         U95
 0    10504397.00000    10447323.00000    7371141.50000    13999447.00000
 1       0.00017       0.00018       0.00003       0.00029
 2       0.00017       0.00018       0.00003       0.00029
 .
 .
 .
 12367       0.00025       0.00021       0.00012     0.00048

From the output, you can see that I used 12367 SNPs, and the positions were given in base pairs not in kilo base pairs.

Thanks in advance.

Answer 1

[GabrielMontenegro]

In case anyone was interested I emailed on of the authors and got a nice reply:

The Mean rho estimate after Loci 0, is the average population recombination rate of my whole region. However this value is WAY too big. He gave me the following advices:

1) To change my SNP positions from bp to kb.

2) If analysing a big (several Mb) region of the human genome it would be best to estimate average rho by sliding windows, say using 2000 SNPs per window.

Answer 2

[nineday.lzu]

I am using LDhat package to estimate recombination/mutation rate with 10kbp windows along genome (Plant). but i found the results is bigger than 1 which should be smaller than 1. the nums after loci 0 of res.txt was treated as recombination rate (map length), and, theta was calculated by formula : theta=S/an. do you have any suggestions for my problem. Thank you very much