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
Pinned
Hanbin Lee
@epigenci.bsky.social
· Jul 19
Modelling complex traits with ancestral recombination graphs https://www.biorxiv.org/content/10.1101/2025.07.14.664631v1
My talk from this year's probgen is finally out.
Where do linear mixed models and random effects come from?
They emerge from *mutations* on Ancestral Recombination Graphs.
Where do linear mixed models and random effects come from?
They emerge from *mutations* on Ancestral Recombination Graphs.
November 26, 2025 at 10:23 PM
I think this is literally true. I kinda recall some of my own rants during the session.
Medical Student Practices Fat Shaming On Cadaver https://theonion.com/medical-student-practices-fat-shaming-on-cadaver/
November 26, 2025 at 9:10 PM
I think this is literally true. I kinda recall some of my own rants during the session.
Is american football just really just an american thing and not popular in rest of North America? I've been to several Mexican restaurants recently and everyone is watching the European Champions League.
November 26, 2025 at 6:10 PM
Is american football just really just an american thing and not popular in rest of North America? I've been to several Mexican restaurants recently and everyone is watching the European Champions League.
Reposted by Hanbin Lee
We are excited to share our recent work on the surprising robustness of Ancestral Recombination Graph (ARG) inference tools to computational phasing errors, now available on BioRxiv: biorxiv.org/content/10.1....
This work is co-advised by @yundeng.bsky.social and Rasmus Nielsen.
1/7
This work is co-advised by @yundeng.bsky.social and Rasmus Nielsen.
1/7
Robustness of Ancestral Recombination Graph Inference Tools to Phasing Errors
Ancestral Recombination Graphs (ARGs) are fundamental population genetic structures that encode the genealogical history of a sample of haplotypes along the genome. They have recently received substan...
biorxiv.org
November 26, 2025 at 5:09 PM
We are excited to share our recent work on the surprising robustness of Ancestral Recombination Graph (ARG) inference tools to computational phasing errors, now available on BioRxiv: biorxiv.org/content/10.1....
This work is co-advised by @yundeng.bsky.social and Rasmus Nielsen.
1/7
This work is co-advised by @yundeng.bsky.social and Rasmus Nielsen.
1/7
Reposted by Hanbin Lee
We've got a really cool preprint out in collaboration with @lpachter.bsky.social's awesome student Cat Felce. Using biophysical models and RNA-seq data, we explore the mechanisms of selective constraint on mRNA abundance, finding constraint on decay rate www.biorxiv.org/content/10.1...
Biophysical constraints on mRNA decay rates shape macroevolutionary divergence in steady-state abundances
Evolutionary changes to gene expression are understood to be a major driver of phenotypic divergence between species. Researchers have investigated the drivers of this divergence by fitting evolutiona...
www.biorxiv.org
November 26, 2025 at 5:25 PM
We've got a really cool preprint out in collaboration with @lpachter.bsky.social's awesome student Cat Felce. Using biophysical models and RNA-seq data, we explore the mechanisms of selective constraint on mRNA abundance, finding constraint on decay rate www.biorxiv.org/content/10.1...
Reposted by Hanbin Lee
When fixed effects aren't fixed...
Important work by @dlmillimet.bsky.social and Marc Bellemare that should influence econometric practice.
onlinelibrary.wiley.com/doi/10.1111/...
Important work by @dlmillimet.bsky.social and Marc Bellemare that should influence econometric practice.
onlinelibrary.wiley.com/doi/10.1111/...
On the (Mis) Use of the Fixed Effects Estimator
Data that span multiple units and time periods allow controlling for time-invariant heterogeneity correlated with the covariates. While researchers can do this in different ways, the fixed effects es...
onlinelibrary.wiley.com
November 26, 2025 at 3:14 PM
When fixed effects aren't fixed...
Important work by @dlmillimet.bsky.social and Marc Bellemare that should influence econometric practice.
onlinelibrary.wiley.com/doi/10.1111/...
Important work by @dlmillimet.bsky.social and Marc Bellemare that should influence econometric practice.
onlinelibrary.wiley.com/doi/10.1111/...
Reposted by Hanbin Lee
I reject the AI booster insistence that you must understand any and everything about the tech to make even the tiniest critique, if they can be vibe coders I can be a vibe hater
November 26, 2025 at 12:33 PM
I reject the AI booster insistence that you must understand any and everything about the tech to make even the tiniest critique, if they can be vibe coders I can be a vibe hater
Reposted by Hanbin Lee
So this happened in Oslo yesterday. Four articulated buses got stuck in a roundabout.
November 25, 2025 at 8:23 AM
So this happened in Oslo yesterday. Four articulated buses got stuck in a roundabout.
Reposted by Hanbin Lee
Just want to sit on that grass in the middle with a couple of beers
So this happened in Oslo yesterday. Four articulated buses got stuck in a roundabout.
November 26, 2025 at 10:38 AM
Just want to sit on that grass in the middle with a couple of beers
Reposted by Hanbin Lee
It's been awhile, but I found myself recommending to someone today our groundbreaking code, lead by the inimitable Andy Kern, that magically converts pdfs to... pdfs. This solves the conversion issue so many journals pose to the intrepid scientist. github.com/andrewkern/p...
GitHub - andrewkern/pdf2pdf: does the magical pdf-to-pdf conversion popular in academic journal submission sites and does it faster than they do
does the magical pdf-to-pdf conversion popular in academic journal submission sites and does it faster than they do - andrewkern/pdf2pdf
github.com
November 26, 2025 at 12:29 AM
It's been awhile, but I found myself recommending to someone today our groundbreaking code, lead by the inimitable Andy Kern, that magically converts pdfs to... pdfs. This solves the conversion issue so many journals pose to the intrepid scientist. github.com/andrewkern/p...
The department coffee often tastes like sesame oil so I searched how close are the two plant species. Not even remotely close.
November 25, 2025 at 7:51 PM
The department coffee often tastes like sesame oil so I searched how close are the two plant species. Not even remotely close.
Doing all these intrinsic calculations feels like being disguised by some evil will.
November 25, 2025 at 6:47 PM
Doing all these intrinsic calculations feels like being disguised by some evil will.
www.biorxiv.org
November 25, 2025 at 3:38 PM
Reposted by Hanbin Lee
Unrelated to mim itself, but this is also the first preprint I've prepared in @typst.app rather than LaTeX. It was soooo much nicer. Folks; what are we doing? Why don't our journals accept manuscript sources in Typst!
Ok; mim (github.com/COMBINE-lab/...) preprint submitted! Excited for folks to see it and share thoughts. The key takeaway; mim allows the quick, one-time, building of a small auxiliary index that then allows scaling gzipped FASTQ parsing linearly in # of threads. 1/2
GitHub - COMBINE-lab/mim: A small, auxiliary index to massively improve parallel fastq parsing
A small, auxiliary index to massively improve parallel fastq parsing - COMBINE-lab/mim
github.com
November 25, 2025 at 2:38 PM
Unrelated to mim itself, but this is also the first preprint I've prepared in @typst.app rather than LaTeX. It was soooo much nicer. Folks; what are we doing? Why don't our journals accept manuscript sources in Typst!
Reposted by Hanbin Lee
Reposted by Hanbin Lee
There is beauty everywhere but especially in autumn in Philadelphia! 😻
November 24, 2025 at 5:32 PM
There is beauty everywhere but especially in autumn in Philadelphia! 😻
Pre-thanksgiving HW sets are a disaster.
November 24, 2025 at 9:36 PM
Pre-thanksgiving HW sets are a disaster.
Reposted by Hanbin Lee
"If you have to explain your sexuality when someone asks if you're dating anyone, you're queer."
February 4, 2024 at 7:15 PM
"If you have to explain your sexuality when someone asks if you're dating anyone, you're queer."
Reposted by Hanbin Lee
“queer not as being about who you’re having sex with (that can be a dimension of it); but queer as being about the self that is at odds with everything around it and has to invent and create and find a place to speak and to thrive and to live.” -bell hooks
February 3, 2024 at 3:21 PM
“queer not as being about who you’re having sex with (that can be a dimension of it); but queer as being about the self that is at odds with everything around it and has to invent and create and find a place to speak and to thrive and to live.” -bell hooks
The causal viewpoint is of course important, but the underlying population genetic setting has to be addressed in advance as it accounts for the ideal setting without causal complications.
1/n
1/n
I didn’t inhibit my argueing muscle enough. this is actually the right take, it’s eventually a case of “what’s your estimand?”. And it’s all observational, though it needs rigorous models before you can get to an answer (which can be, “these aren’t comparable ever”)
SNP h2 is not even h2 and even with we assume all these to be true, it all breaks up when there's selection and drift.
November 24, 2025 at 3:01 PM
The causal viewpoint is of course important, but the underlying population genetic setting has to be addressed in advance as it accounts for the ideal setting without causal complications.
1/n
1/n
Also, SNP heritability being similar across populations isn't even independent evidence because majority of common SNPs driving most of the signal are shared, thereby creating pseudoreplication.
November 24, 2025 at 2:01 AM
Also, SNP heritability being similar across populations isn't even independent evidence because majority of common SNPs driving most of the signal are shared, thereby creating pseudoreplication.
This is so funny.
November 23, 2025 at 8:31 PM
This is so funny.
I just keep reading this piece over and over again and...
arxiv.org/abs/1808.07717
arxiv.org/abs/1808.07717
Multiple Sequence Alignment is not a Solved Problem
Multiple sequence alignment is a basic procedure in molecular biology, and it is often treated as being essentially a solved computational problem. However, this is not so, and here I review the evide...
arxiv.org
November 23, 2025 at 6:09 PM
I just keep reading this piece over and over again and...
arxiv.org/abs/1808.07717
arxiv.org/abs/1808.07717