Specific information on HDS

[bgb-ipl]

Hi,

I need some information regarding the HDS output of Minimac4 (already discussed in the issue #26 ) : could you let me know if the HDS at two different variants can be considered independent ?
For example, let's consider an individual and two variants (denoted SNPa with its 2 HDSs HDSa1/HDSa2 and SNPb with HDSb1/HDSb2). Could HDS be used to compute, for this individual, the dosage of the haplotype with the alternate alleles at both variants (denoted h11) using this formula:

h11 = HDSa1 * HDSb1 + HDSa2 * HDSb2 ?

If anything is unclear don't hesitate to let me know.

Thanks in advance,

[yukt]

You will need to consider LD structure. Variants can be correlated. Best, Ketian

…

On Wed, May 22, 2024 at 9:21 AM bgb-ipl ***@***.***> wrote: Hi, I need some information regarding the HDS output of Minimac4 (already discussed in the issue #26 <#26> ) : could you let me know if the HDS at two different variants can be considered independent ? For example, let's consider an individual and two variants (denoted SNPa with its 2 HDSs HDSa1/HDSa2 and SNPb with HDSb1/HDSb2). Could HDS be used to compute, for this individual, the dosage of the haplotype with the alternate alleles at both variants (denoted h11) using this formula: h11 = HDSa1 * HDSb1 + HDSa2 * HDSb2 ? If anything is unclear don't hesitate to let me know. Thanks in advance, — Reply to this email directly, view it on GitHub <#72>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD6UVLIAE5NLQIE7KNKCTEDZDSLUDAVCNFSM6AAAAABIDUQEEKVHI2DSMVQWIX3LMV43ASLTON2WKOZSGMYTANJSG4YTKOI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[bgb-ipl]

Thank you for your quick answer. Indeed if the HDS do not take the LD into account, the proposed formula is thus not correct.

Best,