Statistical geneticist at Takeda. Views and opinions are my own. Like/retweet != endorsement. genomic.social/@cdeboever3
Joined August 2009
- Tweets 2,424
- Following 1,252
- Followers 453
- Likes 2,855
32 Photos and videos
@cdeboever3@genomic.social retweeted
8 Dec 2022
We predicted 1538 phenotypes for 400K participants in the UK Biobank using the miracle of deep neural networks. Now available at: academic.oup.com/jamia/advan…
26
121
[AS] Melody transcription via generative pre-training
C Donahue, J Thickstun, P Liang [Stanford University] (2022)
arxiv.org/abs/2212.01884
#MachineLearning #ML #AI #Sound
1
5
7
@cdeboever3@genomic.social retweeted
3 Dec 2022
How good is using the closest gene to select the likely causal gene in a GWAS and are there times we can have greater confidence in selecting the closest gene?
I didn't have time yesterday to respond more fully, so I took a look today
@MJ_Schipper, @ksuhre, @DPostu
1
15
56
@cdeboever3@genomic.social retweeted
3 Dec 2022
Machine learning applied to electronic health records identifies four Long-COVID subphenotypes that differ in the involvement of organ systems, previous SARS-CoV-2 infection severity and underlying conditions, according to a @NatureMedicine paper. go.nature.com/3EXZMmx
21
57
@cdeboever3@genomic.social retweeted
1 Dec 2022
Although a recent high-profile paper claimed that synonymous mutations could be just as important as nonsynonymous ones in causing human disease, we argue that synonymous mutations are predominantly neutral in our new @AJHGNews Commentary
authors.elsevier.com/a/1gAt-…
1
34
91
@cdeboever3@genomic.social retweeted
25 Nov 2022
This work from lead by Yi Zeng has found a great home @MolecularCell. If you're interested in branch-points, timing of splicing, and order of splicing take a look. cell.com/molecular-cell/full…
23
84
@cdeboever3@genomic.social retweeted
24 Nov 2022
1/n Single-cell data analysis is complex with many tools to choose from - but which ones work best and should be used when? Led by @lukasheumos and @AnnaCSchaar, we wrote a free online book that aims to guide single-cell data analysts: sc-best-practices.org
15
403
1,046
@cdeboever3@genomic.social retweeted
22 Nov 2022
New #proteomcis GWAS with @SomaLogic with largest SOMAScan 4.1 platform to date: Identification of 969 protein quantitative trait loci in an African American population with kidney disease attributed to hypertension kidney-international.org/art…
2
7
18
@cdeboever3@genomic.social retweeted
20 Nov 2022
What do these 3 papers have in common?
pubmed.ncbi.nlm.nih.gov/3023…
pubmed.ncbi.nlm.nih.gov/3534…
pubmed.ncbi.nlm.nih.gov/3635…
5
31
194
@cdeboever3@genomic.social retweeted
17 Nov 2022
Our manuscript describing the genetic constraint of splicing in coding and non-coding regions of proteins-coding genes is out! We use this model to predict deleterious splicing at a per-base resolution with ConSpliceML in #RareDisease. #Bioinformatics #ML bmcbioinformatics.biomedcent…
9
38
@cdeboever3@genomic.social retweeted
16 Nov 2022
I am so excited to share my first first-author paper, out today @Nature!!
So thankful to work with three other amazing first-authors @JoelWBlanchard1 @_josedavila and Djuna von Maydell, and an inspiring team lead by @DrLiHueiTsai 🤩🔥
nature.com/articles/s41586-0…
22
84
581
Today a 120B model called “Galactica” is open-sourced by @paperswithcode. It’s capable of writing math notations, citations, code, chemical formula, DNA, etc. Here’s why I think Galactica is a huge milestone in open foundation models, scientific automation, and responsible AI: 🧵
29
330
1,734
@cdeboever3@genomic.social retweeted
16 Nov 2022
Nova’s “Zero to Infinity” hosted by Talithia Williams @Dr_TalithiaW, premieres on @PBS Wednesday night 9pmET and streams on Nova’s website and YouTube channels earlier in the day:
pbs.org/wgbh/nova/video/zero…
3
11
@cdeboever3@genomic.social retweeted
15 Nov 2022
Identification of PCSK9-like human gene knockouts using metabolomics, proteomics, and whole-genome sequencing in a consanguineous population cell.com/cell-genomics/fullt…
5
26
80
@cdeboever3@genomic.social retweeted
10 Nov 2022
metaphor.systems is now publicly available!
Metaphor is a search engine based on generative AI, the same sorts of techniques behind DALL-E 2 and GPT-3
1/
73
542
2,733
@cdeboever3@genomic.social retweeted
10 Nov 2022
This is now out.
Reminder: Study only compares people with multiple infections to those without, not reinfections to first infections.
"If you fell twice, are you worse off compared with person who only fell once, especially if falls correlate with frailty in the first place".
23 Jun 2022
Study design of that reinfection preprint below.
Day one: Alice and Bob are infected.
Day, say, 45: Bob is reinfected. Alice is never reinfected.
Six months later, Alice is in better health than Bob. That's *all* the study measures.
Not comparing first vs further infections.
28
112
425
@cdeboever3@genomic.social retweeted
10 Nov 2022
New report from Germany of a large matched-control #LongCovid study of 157,000 people w/Covid, including ~12,000 children and adolescents
"The results of the present study indicate that post-Covid-19 cannot be dismissed among children and adolescents."
journals.plos.org/plosmedici…
32
631
1,328
@cdeboever3@genomic.social retweeted
10 Nov 2022
Covid reinfections are bad for your health. Big @DeptVetAffairs study published @NatureMedicine today. #LongCovid
nature.com/articles/s41591-0…
by @zalaly and colleagues
88
1,453
3,029
@cdeboever3@genomic.social retweeted
8 Nov 2022
What do polygenic risk scores (PRS) add to the seemingly simple question about family history (FH)?
We explored the interplay of PRS and FH in over 20 common diseases, showing that they have little overlap and provide complementary information.
Link: bit.ly/fhprs
4
98
279
@cdeboever3@genomic.social retweeted
20 Mar 2022
Happy to share our new preprint on meta-analysis fine-mapping in the Global Biobank Meta-analysis Initiative @GlobalBiobanks. Joint work w/ Roy Elzur @weizhouw under the mentorship of @dalygene Hilary Finucane. Tl;dr meta-analysis fine-mapping is hard! 1/n
medrxiv.org/content/10.1101/…
2
26
76