Christa Caggiano

Extremely happy to see the last paper of my PhD published, which focused on using cell-free DNA to detect and monitor ALS worked with an amazing cross-continent team, which was co-led by Noah Zaitlen and @FleurryG link.springer.com/article/10.118…

This is the last week to take our survey! If you're an early career researcher in human genetics and have some time, please consider taking our survey on how you use population descriptors It's a study run by trainees with the goal of improving trainee research quality!

Are you an early-career researcher working on the computational analysis of population-level human genetics data? We want to hear from you about if, how, and why you use population descriptors in your research! Fill out our short survey: forms.gle/SCiNUq71wgi5co…

If you are a trainee in genetics and have 10 minutes of time today/feel like procrastinating, please take our survey. This project is completely trainee led, and we want to get as broad of a perspective on how trainees are doing this research as possible!

@nunorgcarvalho @roohys @SineadCullina @popgenepi @EimearEKenny Wow such a nice thread! :)

Some of those authors on this pre-print include: @roohys @SineadCullina @ChristaCaggiano @popgenepi @EimearEKenny

They then build an adjacency matrix, which for any pair of individuals, assigns them a 1 if they share more than 6cM of IBD, and 0 otherwise. It's essentially a matrix of relatives.

A powerful alternative to principal components (PC) to account for population structure in genetic association analysis: spectral components (SPc) Compared to PC, SPc - explain more variation in population structure with fewer components (90% for SPc vs 50% for PC) - more effectively reduce inflation of P values from GWAS of environmental-driven phenotypes - improve rare variant association analysis - help calculate more accurate heritability estimates Reference (link in comment): Shemirani et al. SPC: a SPectral Component approach to address recent population structure in genomic analysis. medRxiv 2025

@CharlestonCWKC Strongly agree! I think when it’s all results no context it’s very difficult to read if you’re not in that exact subfield

This might be personal style. I prefer Results section include a (brief) discussion of results while Discussion section as broader implications. But there are people who write the Results section as purely an objective statement of results (no motivation, no conclusion).

Hi academic friends! Does anyone have a recent-ish NIH F32 that they would be willing to share? I can’t find any online written within 10 years, but I’d be super appreciative if anyone has an example 🙂

@grocklin I still am a little salty about reading this in depth for my boot camp and then no one ever asking about it 😅

Saw a piece of paper sticking out of Schrödinger's "What is Life?" on my bookshelf. Turns out it was discussion questions for UCSF's incoming biophysics student bootcamp in 2011. They were all supposed to read the book.

Check out our new toolkit for cell type deconvolution from DNA methylation sequencing. Open-source and ready to use! Congrats to @UntIrene who worked tirelessly on this & thanks to @ChristaCaggiano and Noah Zaitlen who inspired us with the CelFiE approach. genomebiology.biomedcentral.com/articles/10.11…

@benbfly @UntIrene Yay so happy to see this out!!

How do quantities of brain cells – the fundamental unit of life – differ in Alzheimer’s disease, schizophrenia and autism? Our work on cell-type proportion shifts using methylomics is just out in @ScienceAdvances #PsychENCODE @mikejg84 @DanielVo10 science.org/doi/10.1126/sc… 🧵

@_henry_snow UCLA has access to this database but not this specific article 🥲

So happy to have been at the beautiful UPitt for the inaugural #statgen2024. Super fun conference with lots of cool stat gen methods, definitely a new fave

@hjpimentel Big mistake making the office hours at a time grad students are awake. Next year make the office hours at 8am

Please send thoughts and prayers: homework for 200 students due tonight and my office hour will begin soon…

Our paper looking at cell-free DNA epigenetic features in ALS is now out in MedRxiv! I am extraordinarily grateful for the ALS patients who participated in the study. This research would not be possible without them medrxiv.org/content/10.110…

Really wonderful work on how our genomes might interact with commonly prescribed medications

I noticed ppl disregarding the NYMag piece bc it focused on Huberman’s personal life, not bad science & harmful misinformation he propagates. So I wrote a thing. @slate: “So, Should You Trust Andrew Huberman?” slate.com/technology/202…

@Ugo_alves The tldr is that there’s really variable PGS quality, these guys are just mining public ones. At best, I’m worried people will make irrevocable decisions about their health or future children’s health based on basically no evidence. At worst, it could be eugenics-y

@Ugo_alves Then lastly, human diseases often share genetic backgrounds. A depression score might be predictive because it shares key mutations with bipolar disorder and there’s correlation between those diseases. Knowing your PGS for depression then doesnt necessarily help w good treatment

Nothing makes me as uneasy as a human geneticist as seeing the proliferation of direct to consumer polygenic risk score companies. I know academics are conservative but these scores are truly not good enough yet & have potential for true harms