Nils Wagner
@nils_f_wagner
Followers
60
Following
3
Media
13
Statuses
31
Is binarization of scATAC-seq data really necessary? The conclusion from our analysis is that a quantitative treatment is in fact beneficial. Now out in Nature Methods! @gagneurlab @fabian_theis
https://t.co/UdFyU8fpYz Many additions since the preprint 👇(1/n)
4
97
320
Looking forward to David Kelley's talk on Wednesday at the Kipoi seminar presenting Borzoi https://t.co/smqvKYDnAk
0
3
15
In which tissues a variant may disrupt splicing? AbSplice predicts this, also allowing leveraging RNA-seq of your patients from clinically accessible tissues, eg blood or fibroblast. Come to my poster PB1021 at #ASHG23 today 3pm to learn more & chat about application on your data
Very excited to announce that our study on aberrant splicing prediction across human tissues is now out @NatureGenet
https://t.co/sNxKR7UULz. Huge thanks to amazing colleagues: @muhammed_hasan, Florian,@chr_mertes, @ProkischLab ,@vaym88, @gagneurlab! (1/7)
0
2
11
Cracking the regulatory code: We have genomes for 1000s of species, but ENCODE only for 2 – what do we do? Natural language models have shown that syntax and semantics can be learned from text alone. Can we do the same for genomes?⬇️ https://t.co/MIFqNVfAlC
biorxiv.org
The rise of large-scale multi-species genome sequencing projects promises to shed new light on how genomes encode gene regulatory instructions. To this end, new algorithms are needed that can...
3
31
119
Press release about our aberrant splicing prediction tool AbSplice.
Researchers developed a new algorithm that is six times more reliable than previous models at predicting the effects of #geneticmutations on #RNA formation that may cause rare #hereditarydiseases or #cancer: https://t.co/GI59Kzc3HO
@gagneurlab 📷D.Gankin
0
2
15
Researchers developed a new algorithm that is six times more reliable than previous models at predicting the effects of #geneticmutations on #RNA formation that may cause rare #hereditarydiseases or #cancer: https://t.co/GI59Kzc3HO
@gagneurlab 📷D.Gankin
1
5
19
🎉 OUT TODAY @NatureGenet 📰 Aberrant splicing prediction across human tissues 👩🏻🤝👨🏼 Nils Wagner, Julien Gagneur, and colleagues 👇🏿 https://t.co/hBuh0hkaKC
0
18
44
Thanks to my colleague @muhammed_hasan for wonderful teamwork and the whole team @gagneurlab!
1
0
1
For more check out the published article and don’t hesitate to contact us! Precomputed AbSplice scores for all possible SNVs genome-wide and 49 GTEx tissues are available at https://t.co/OaqqSazt2s. The AbSplice code is publicly available at https://t.co/JfVY02elKp. (7/7)
github.com
Contribute to gagneurlab/absplice development by creating an account on GitHub.
1
0
2
We compared different outlier callers (LeafcutterMD, SPOT, FRASER) and added additional baseline models (CADD-Splice, SQUIRLS, MTSplice) to the benchmark. FRASER led to the best performance and including other baseline models to AbSplice showed only minor improvements. (6/7)
1
0
2
We validated model performances using independent proteomics data from an ALS dataset and found that validation rates were consistent with performances we observed in the GTEx benchmark dataset. (5/7)
1
0
1
We stratified model performances for different outlier outcome types. AbSplice-DNA performed better for exon skipping than for exon elongation and truncation, as well as better for alternative donor or acceptor choice than for splicing efficiency outliers. (4/7)
1
0
2
... we stratified model performances for different variant categories. As expected, the precision was the best on variants affecting the donor and acceptor dinucleotides, followed by variants in the splice region, then in the exonic, and lastly in the intronic regions. (3/7)
1
0
2
Additionally to the preprint: https://t.co/C8aGt2RWVG ... (2/7)
Excited to share AbSplice to predict aberrant splicing across tissues. We created the 1st benchmark for aberrant splicing, improved DNA-based predictors with a tissue-specific splicing map and showed how to integrate RNAseq of accessible tissues:
1
0
1
Very excited to announce that our study on aberrant splicing prediction across human tissues is now out @NatureGenet
https://t.co/sNxKR7UULz. Huge thanks to amazing colleagues: @muhammed_hasan, Florian,@chr_mertes, @ProkischLab ,@vaym88, @gagneurlab! (1/7)
nature.com
Nature Genetics - AbSplice predicts aberrant splicing for 50 human tissues by integrating sequence-based deep learning models, DNA variation and RNA-seq obtained from accessible tissues.
3
41
161
Proud to announce that our paper on deep models of gene expression is now out @GenomeBiology
https://t.co/xR7Cn17q8H. In addition to extensive benchmarks in the preprint (see RT)... https://t.co/0HJ4jkM9DU (1/6)
Xpresso, Basenji, Enformer,... Where do sequence-based models of transcription stand? In our new preprint, Alex Karollus tested SOTA models against 2 RNA-seq studies and 5 deep perturbation assays testing promoters, enhancers, eQTL 1/n https://t.co/frpZSAoNMb
2
14
51
Xpresso, Basenji, Enformer,... Where do sequence-based models of transcription stand? In our new preprint, Alex Karollus tested SOTA models against 2 RNA-seq studies and 5 deep perturbation assays testing promoters, enhancers, eQTL 1/n https://t.co/frpZSAoNMb
biorxiv.org
Background The largest sequence-based models of transcription control to date have been obtained by predicting genome-wide gene regulatory assays across the human genome. This setting is fundamenta...
3
58
170
Poster PB2506 "Aberrant splicing prediction across human tissues", Thu 3pm - 4:45pm #ASHG22. Preprint:
0
5
7