Finding - Single Parent

[CGWallace]

Eric:

What are you plans with SNPpit – in particular single parent calls. I have been using Colony but now that we are getting into Chinook the baselines as well as the mixtures are getting large and therefore can take a lot of time to run. The single parent issue is the only thing stopping us from using SNPpit.

Colin Wallace Molecular Genetics Lab

[eriqande]

Hi Colin,

Unfortunately, I don't see that digging back into the guts of SNPPIT and doing single parent assignment as the best option. I do have an R package called CKRMsim that does single parent stuff pretty quickly. My intention has been to clean up the interface for that for single parent stuff. With that, it might be possible to do an interesting hybrid thing:

1. look for parent pairs with SNPPIT
2. Identify single parents with CKMRsim

Use the overlap between the two to calibrate the single parent assignment confidences, the idea being parent-pair assignments are quite confident, and could possibly be used to calibrate the method for the (more difficult) single parent assignments (especially when there are siblings and aunts and uncles in the mix).

So, I will leave this open, but it is more likely you will see development on CKMRsim. I have to start crackin on that anyway for a workshop this summer.

I hope you guys are doing great up there! Big hello to Terry!

eric

[CGWallace]

Thanks again Eric. I will start looking into CKRMsim.

Colin Wallace Molecular Genetics Lab

From: Eric C. Anderson notifications@github.com Sent: February 5, 2019 2:14 PM To: eriqande/snppit snppit@noreply.github.com Cc: Wallace, Colin Colin.Wallace@dfo-mpo.gc.ca; Author author@noreply.github.com Subject: Re: [eriqande/snppit] Finding - Single Parent (#3)

Hi Colin,

Unfortunately, I don't see that digging back into the guts of SNPPIT and doing single parent assignment as the best option. I do have an R package called CKRMsim that does single parent stuff pretty quickly. My intention has been to clean up the interface for that for single parent stuff. With that, it might be possible to do an interesting hybrid thing:

1. look for parent pairs with SNPPIT
2. Identify single parents with CKMRsim

Use the overlap between the two to calibrate the single parent assignment confidences, the idea being parent-pair assignments are quite confident, and could possibly be used to calibrate the method for the (more difficult) single parent assignments (especially when there are siblings and aunts and uncles in the mix).

So, I will leave this open, but it is more likely you will see development on CKMRsim. I have to start crackin on that anyway for a workshop this summer.

I hope you guys are doing great up there! Big hello to Terry!

eric

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/eriqande/snppit/issues/3#issuecomment-460824867, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Af8e66ZjHv4nUzs4QN1V_nXv1UCWTeh8ks5vKgIMgaJpZM4aj_XK.

[eriqande]

Cool.

The documentation for CKMRsim is pretty sparse. Hopefully I will get a day in the next month to get back into the code and clean it up a bit!

eric

On Tue, Feb 5, 2019 at 3:25 PM CGWallace notifications@github.com wrote:

Thanks again Eric. I will start looking into CKRMsim.

Colin Wallace Molecular Genetics Lab

From: Eric C. Anderson notifications@github.com Sent: February 5, 2019 2:14 PM To: eriqande/snppit snppit@noreply.github.com Cc: Wallace, Colin Colin.Wallace@dfo-mpo.gc.ca; Author < author@noreply.github.com> Subject: Re: [eriqande/snppit] Finding - Single Parent (#3)

Hi Colin,

Unfortunately, I don't see that digging back into the guts of SNPPIT and doing single parent assignment as the best option. I do have an R package called CKRMsim that does single parent stuff pretty quickly. My intention has been to clean up the interface for that for single parent stuff. With that, it might be possible to do an interesting hybrid thing:

1. look for parent pairs with SNPPIT
2. Identify single parents with CKMRsim

Use the overlap between the two to calibrate the single parent assignment confidences, the idea being parent-pair assignments are quite confident, and could possibly be used to calibrate the method for the (more difficult) single parent assignments (especially when there are siblings and aunts and uncles in the mix).

So, I will leave this open, but it is more likely you will see development on CKMRsim. I have to start crackin on that anyway for a workshop this summer.

I hope you guys are doing great up there! Big hello to Terry!

eric

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub< https://github.com/eriqande/snppit/issues/3#issuecomment-460824867>, or mute the thread< https://github.com/notifications/unsubscribe-auth/Af8e66ZjHv4nUzs4QN1V_nXv1UCWTeh8ks5vKgIMgaJpZM4aj_XK>.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/eriqande/snppit/issues/3#issuecomment-460828637, or mute the thread https://github.com/notifications/unsubscribe-auth/AB8OJtUbWvtXnd-1pjSaH23OcmV_ZjVcks5vKgTsgaJpZM4aj_XK .

[eriqande]

Here is a tutorial on CKMRsim that might be helpful:

https://eriqande.github.io/tws-ckmr-2022/