Hanbin Lee
@epigenci.bsky.social
PhD Student at UMich Statistics.
The account mostly trashes about urban planning and infrastructure.
Probability, Statistics, and Evolutionary Biology.
https://hanbin973.github.io
The account mostly trashes about urban planning and infrastructure.
Probability, Statistics, and Evolutionary Biology.
https://hanbin973.github.io
"Nothing in biology makes sense except in the light of evolution."
well... but only when the light is shed properly.
well... but only when the light is shed properly.
November 9, 2025 at 3:11 PM
"Nothing in biology makes sense except in the light of evolution."
well... but only when the light is shed properly.
well... but only when the light is shed properly.
Reposted by Hanbin Lee
Watson's misogynistic and racist comments not only embarrassed Watson and those associated with him, they also gave license to others to express similar views about the genetic inferiority of different groups of human beings and even acceptable steps for dealing this this.
10/13
10/13
November 9, 2025 at 1:33 AM
Watson's misogynistic and racist comments not only embarrassed Watson and those associated with him, they also gave license to others to express similar views about the genetic inferiority of different groups of human beings and even acceptable steps for dealing this this.
10/13
10/13
kind of funny that this came out after the election, which is the most puzzling part (was the decision made before the election tho?)
November 8, 2025 at 2:29 PM
kind of funny that this came out after the election, which is the most puzzling part (was the decision made before the election tho?)
The only caveat is that it needs a lot of resting/reheating which takes a lot of time. I save time by heating up the meat while I cook something else for lunch.
November 7, 2025 at 1:18 PM
The only caveat is that it needs a lot of resting/reheating which takes a lot of time. I save time by heating up the meat while I cook something else for lunch.
Reposted by Hanbin Lee
@kjgilbert.bsky.social [https://kjgilbert.github.io/research/] has you covered:
November 4, 2025 at 6:50 PM
@kjgilbert.bsky.social [https://kjgilbert.github.io/research/] has you covered:
This tells us that population structure itself, which does not contribute to "directional" allele frequency stratification, can't create fixed effects. What happens there is selection? well that I think is still an open question.
(8/8).
(8/8).
November 5, 2025 at 10:58 PM
This tells us that population structure itself, which does not contribute to "directional" allele frequency stratification, can't create fixed effects. What happens there is selection? well that I think is still an open question.
(8/8).
(8/8).
This makes PCs obsolete as a means to handle fixed genetic contributors. Of course, there are non-genetic contributors aligning with genetic factors so PCs might have role, but it has nothing to do with genetics.
(7/n)
(7/n)
November 5, 2025 at 10:56 PM
This makes PCs obsolete as a means to handle fixed genetic contributors. Of course, there are non-genetic contributors aligning with genetic factors so PCs might have role, but it has nothing to do with genetics.
(7/n)
(7/n)
As such, any fixed contribution of genetics to individuals are constant. Of course, the random portion covaries according to population structure but this is accounted by the random effects. That's why we have off-diagonal elements in the GRM. So the fixed effect is merely a constant.
(6/n)
(6/n)
November 5, 2025 at 10:54 PM
As such, any fixed contribution of genetics to individuals are constant. Of course, the random portion covaries according to population structure but this is accounted by the random effects. That's why we have off-diagonal elements in the GRM. So the fixed effect is merely a constant.
(6/n)
(6/n)
This means that from the founder or the root (of the genealogy) to the present, the expected amount of mutations (and subsequently the genetic values) that haplotypes accumulate over time is constant for everyone.
(5/n)
(5/n)
November 5, 2025 at 10:51 PM
This means that from the founder or the root (of the genealogy) to the present, the expected amount of mutations (and subsequently the genetic values) that haplotypes accumulate over time is constant for everyone.
(5/n)
(5/n)
This is because both are treating mutations as random AND the evolution is assumed to be neutral although sophisticated demography and population structure are allowed.
(4/n)
(4/n)
November 5, 2025 at 10:50 PM
This is because both are treating mutations as random AND the evolution is assumed to be neutral although sophisticated demography and population structure are allowed.
(4/n)
(4/n)
Hence, the two are not strictly the same. Although the inference is applied to the same data, the former is fitting P(data|pedigree) while the latter is fitting P(data|genealogy).
Nevertheless, they arrive at the same conclusion. but why?
(3/n)
Nevertheless, they arrive at the same conclusion. but why?
(3/n)
November 5, 2025 at 10:48 PM
Hence, the two are not strictly the same. Although the inference is applied to the same data, the former is fitting P(data|pedigree) while the latter is fitting P(data|genealogy).
Nevertheless, they arrive at the same conclusion. but why?
(3/n)
Nevertheless, they arrive at the same conclusion. but why?
(3/n)