It is written in the manual that the errors in marker orders can be detected by making a pair-wise LOD plot of the resulting markers and that the pair-wise data can be generated by LM2.
Would it be possible to give some comment on how to generate the pair-wise LOD data by LM2?
Thanks,
Homa
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I followed your instructions and I could generate the following plot. I have 21 families with a total of 230 individuals so I set the LOD to 3 in OrderMarkers. I am creating the linkage map only for the sex chromosome. In the plot, I don't see the brighter squares overlapping the diagonal indicating of linkage of close markers in contrast the clear pattern in your graph.
It would be a great help for me if you could let me know what is your opinion on the attached plot? Does it mean that marker orders may not be correct? Although after running OrderMarkers 20 times, I have a convergence in likelihood and the overall marker orders.
Plot looks quite good. It would be easier to see if you would cap the LOD scores, say at 30 (In gnuplot, set zrange [0:30]).
There is a bit strange part in markes 275-300. As there are many families, the stripiness of the plot may be caused by difference in the number of informative families of each marker. The markers positions will tell whether these markers are all put to same position or not.
I will attach here a plot with first 100 markers inverted.
Dear Pasi,
It is written in the manual that the errors in marker orders can be detected by making a pair-wise LOD plot of the resulting markers and that the pair-wise data can be generated by LM2.
Would it be possible to give some comment on how to generate the pair-wise LOD data by LM2?
Thanks,
Homa
Dear Homa,
Thank you for your question.
Yes this is possible. With the simulated data I made the following plot this way:
java -cp bin OrderMarkers data=data1.linkage evaluateOrder=order.txt improveOrder=0 computeLODScores=1 >lod1.txt
And the image was created in gnuplot as follows:
set terminal png size 2000, 2000
set output "lod1.png"
set pm3d map
splot "< cut -f 2- lod1.txt|awk '(NR>2)'" matrix
Cheers,
Pasi
Last edit: Pasi Rastas 2016-08-31
Dear Pasi,
Thank you very much for your reply.
I followed your instructions and I could generate the following plot. I have 21 families with a total of 230 individuals so I set the LOD to 3 in OrderMarkers. I am creating the linkage map only for the sex chromosome. In the plot, I don't see the brighter squares overlapping the diagonal indicating of linkage of close markers in contrast the clear pattern in your graph.
It would be a great help for me if you could let me know what is your opinion on the attached plot? Does it mean that marker orders may not be correct? Although after running OrderMarkers 20 times, I have a convergence in likelihood and the overall marker orders.
Thank you very much for your help,
Cheers,
Homa
Last edit: Homa 2016-09-01
Dear Homa,
Plot looks quite good. It would be easier to see if you would cap the LOD scores, say at 30 (In gnuplot, set zrange [0:30]).
There is a bit strange part in markes 275-300. As there are many families, the stripiness of the plot may be caused by difference in the number of informative families of each marker. The markers positions will tell whether these markers are all put to same position or not.
I will attach here a plot with first 100 markers inverted.
Cheers,
Pasi