Markers from different chromosomes on the same LG

[Ana Mota]

Hello!

I am very new in this type of analysis (make linkage maps), and I am working with two different populations (F2). The species i am working on has a reference genome well stablished, and I know the number of chromosomes expected.
The problem I'm facing is :
I'm using the filtered markers as input for lepmap3.
I'm following the different pipelines to produce the linkage map as follow:
1- lepmap3 ParentCall2 data=pedigree.2.txt vcfFile=pop_1_maf03_miss_08_bi.vcf.recode.vcf > p.call
2- lepmap3 Filtering2 data=p.call removeNonInformative=1 dataTolerance=0.0000001 > p_fill.call
3- lepmap3 SeparateChromosomes2 data=p_fill.call lodLimit=13 sizeLimit=15 distortionLod=1 numThreads=20 > map.txt
This first map is giving me almost the number of chromosomes expected (one chromosome less than expected). I have tried several different parameters for this step, and this one is the one giving me the best.

4- lepmap3 OrderMarkers2 data=p_fill.call map=map.txt chromosome=1 computeLODScores=lod_scores calculateIntervals=int_calcul numThreads=20 > order.txt

Even if I have a very close number of chromosomes in this analysis, when I check the markers, some LG are composed by markers from different chromosomes.
My question is: there is a way to use the reference genome to prepare the genetic map ? Can I divide my input file per chromosome, and try to do the analysis per chromosome ?
Also, do you recommend extra steps for filtering the markers on the VCF file? Like segregation distortion?

Thank you very much,
Ana

[Pasi Rastas]

Dear Ana,

Thank you for your question.

Can you share the number of marker in each linkage group? Typically you have several small groups which can be ignored. Below is the unix command to obtain this:

sort map.txt|uniq -c|sort -n

You hope to have the expected number of large groups and several much smaller ones. You seem to remove groups smaller than 15 with sizeLimit.

It is very rare that you have to filter VCF files, typically Lep-MAP3 can do required filtering.
For more aggressive segregation distortion filtering, increase the parameter dataTolerance=0.0000001 (e.g. to 0.00001).

Cheers,
Pasi