Pontus Skoglund
@pontus-skoglund.bsky.social
Ancient DNA, evolution, and the human past. Senior group leader leading the research of the Ancient Genomics lab at the Francis Crick Institute. http://www.skoglundlab.org
Reposted by Pontus Skoglund
I’m not an expert on polygenic scores for mental health, but a concern I would have is whether something has gone wrong with the genotyping, leading to misestimated polygenic scores that look extreme.
November 15, 2025 at 3:19 AM
I’m not an expert on polygenic scores for mental health, but a concern I would have is whether something has gone wrong with the genotyping, leading to misestimated polygenic scores that look extreme.
Reposted by Pontus Skoglund
In particular, the claims that “[Hitler’s polygenic] score for schizophrenia, autism and bipolar disorder was actually higher than 99% of the individuals.” and " It is also rare to have a high polygenic risk score for all three conditions" strike me as strange.
November 15, 2025 at 3:19 AM
In particular, the claims that “[Hitler’s polygenic] score for schizophrenia, autism and bipolar disorder was actually higher than 99% of the individuals.” and " It is also rare to have a high polygenic risk score for all three conditions" strike me as strange.
Reposted by Pontus Skoglund
I think we can say confidently that there is zero scientific value in such 'celebrity' genomes, notorious or otherwise, and probably little or no historical value either.
November 15, 2025 at 12:12 AM
I think we can say confidently that there is zero scientific value in such 'celebrity' genomes, notorious or otherwise, and probably little or no historical value either.
Seems not be controlled with estimated PRSs for similarly lower-quality genomes. For 80-year old blood from a sofa, what is the X-fold coverage and genotyping strategy?
If there is a higher error rate than the modern panel, then the scores for rare-variant-driven traits like this could be inflated.
If there is a higher error rate than the modern panel, then the scores for rare-variant-driven traits like this could be inflated.
November 14, 2025 at 12:05 PM
Seems not be controlled with estimated PRSs for similarly lower-quality genomes. For 80-year old blood from a sofa, what is the X-fold coverage and genotyping strategy?
If there is a higher error rate than the modern panel, then the scores for rare-variant-driven traits like this could be inflated.
If there is a higher error rate than the modern panel, then the scores for rare-variant-driven traits like this could be inflated.
Ok thanks Adam...
November 13, 2025 at 11:32 AM
Ok thanks Adam...
Where is the technical information available?
November 13, 2025 at 10:02 AM
Where is the technical information available?