I'm about to try LM2 for a fairly complex pedigree of a natural population and I'm a bit unclear whether the design would work fine for the program. I have ~280 bird samples typed for ~60K SNPs but the system I work on is one of the most promiscuous systems and therefore have high extra-pair paternity rates. There are plenty of half siblings but we do have the genotypes for both parents even when they are half. With that said, there are few samples where we only have data from one parent.
Would this still work? Does LM3 offer an advantage with this type of system? If I have half-sibs but data from both parents, should I code the input for both parents? As the relationships between individuals resemble a complex network, do I define a family based on the network and therefore allele sharing? Let me know if you need more information to help advise whether LM2/3 would work this system.
Thanks in advance!
Best,
Josh
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
PS. When I say complicated, I mean it looks like this. Where the dotted lines link duplicated individuals that needed to be placed in multiple parts of the tree and blacked symbols are individuals for which we don't have data for. All individuals are technically related one way or another, so I'm unclear as to how I would define a family.
Thank you for sharing your pedigree. It looks a bit hairy to me:)
However, there seems to quite many families with enough offspring. The ParentCall2 module in Lep-MAP3 has now half-sib parameter (halfSibs=1). You could try it out. The pedigree has to be split into full-sib families, ParentCall2 will find out indentical individual names to figure out half-sibs. (Warning, this is still in it's preliminary stage). I would take only families with minimum of 2-3 offpspring. A smaller family might be sometimes useful if the same parents have also larger families. However, for constructing maps, 2 is the absolute minimum and 3+ is preferred. More offspring are required if both parents are not genotyped.
Also note that in your pedigree, you might have grandparents as well. You could use them with ParentCall2 as well. Then you could also phase the data and maybe use better the small families... Lep-MAP3 can used phased data (all phased or all unphased).
Cheers,
Pasi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Rasi,
I'm about to try LM2 for a fairly complex pedigree of a natural population and I'm a bit unclear whether the design would work fine for the program. I have ~280 bird samples typed for ~60K SNPs but the system I work on is one of the most promiscuous systems and therefore have high extra-pair paternity rates. There are plenty of half siblings but we do have the genotypes for both parents even when they are half. With that said, there are few samples where we only have data from one parent.
Would this still work? Does LM3 offer an advantage with this type of system? If I have half-sibs but data from both parents, should I code the input for both parents? As the relationships between individuals resemble a complex network, do I define a family based on the network and therefore allele sharing? Let me know if you need more information to help advise whether LM2/3 would work this system.
Thanks in advance!
Best,
Josh
PS. When I say complicated, I mean it looks like this. Where the dotted lines link duplicated individuals that needed to be placed in multiple parts of the tree and blacked symbols are individuals for which we don't have data for. All individuals are technically related one way or another, so I'm unclear as to how I would define a family.
Dear Joshua,
Thank you for sharing your pedigree. It looks a bit hairy to me:)
However, there seems to quite many families with enough offspring. The ParentCall2 module in Lep-MAP3 has now half-sib parameter (halfSibs=1). You could try it out. The pedigree has to be split into full-sib families, ParentCall2 will find out indentical individual names to figure out half-sibs. (Warning, this is still in it's preliminary stage). I would take only families with minimum of 2-3 offpspring. A smaller family might be sometimes useful if the same parents have also larger families. However, for constructing maps, 2 is the absolute minimum and 3+ is preferred. More offspring are required if both parents are not genotyped.
Also note that in your pedigree, you might have grandparents as well. You could use them with ParentCall2 as well. Then you could also phase the data and maybe use better the small families... Lep-MAP3 can used phased data (all phased or all unphased).
Cheers,
Pasi
Hi Pasi,
Thanks for the response. Glad to hear that Lep-MAP3 can handle half-sibs! I'll try it out but I'm sure I'll have more questions as I progres. Thanks!
Best,
Josh