Johannes Hingerl

Excited to share Nona: a unifying multimodal masking framework for functional genomics.

Models for DNA have evolved along separate paths: sequence-to-function (AlphaGenome), language models (Evo2), and generative models (DDSM).

Can these be unified under a single paradigm? 1/15

gReLU: a comprehensive framework for DNA sequence modeling and design - Nature Methods gReLU advances deep-learning-based modeling and analysis of DNA sequences with comprehensive toolsets and versatile applications.

gReLU advances deep learning based modeling and analysis of DNA sequences with comprehensive toolsets and versatile applications. @avantikalal.bsky.social @gokcen.bsky.social

www.nature.com/articles/s41...

Excited to share UKBBGym at #ASHG25, a new benchmark for variant effect predictors using WGS, proteomics and phenotypes from 500K UKBiobank participants. Stop by for insights on the impact of non-coding variants and how computational scores stack up against exp assays.
Poster 5022W, Wed 2:30-4:30.

GitHub - johahi/borzoi-pytorch: Pytorch implementation of the Borzoi model from Calico, and Flashzoi, a 3x faster Borzoi enhancement. Pytorch implementation of the Borzoi model from Calico, and Flashzoi, a 3x faster Borzoi enhancement. - johahi/borzoi-pytorch

Big thanks to co-authors @karollus.bsky.social & @gagneurlab.bsky.social, and the Borzoi authors Johannes Linder, @drkbio.bsky.social et al!
Flashzoi is available on github and on pip: github.com/johahi/borzo...

The speedup does not come at the cost of accuracy:
Flashzoi matches or improves upon Borzoi’s performance across benchmarks, including RNA-seq coverage prediction, variant effect prediction (GTEx eQTLs), and enhancer-gene linking.

Flashzoi: an enhanced Borzoi for accelerated genomic analysis AbstractMotivation. Accurately predicting how DNA sequence drives gene regulation and how genetic variants alter gene expression is a central challenge in

Happy to share that Flashzoi is now published!
We enhanced Borzoi with RoPE & FlashAttention for >3x faster training/inference & 2.4x reduction in memory usage.
This brings large-scale genomic analysis and fine-tuning within reach of academic budgets.
📄: doi.org/10.1093/bioi...

The Biodiversity Cell Atlas white paper is out! A bold vision to map the diversity and evolution of cell types across the tree of life 🌍

We are excited to share GPN-Star, a cost-effective, biologically grounded genomic language modeling framework that achieves state-of-the-art performance across a wide range of variant effect prediction tasks relevant to human genetics.
www.biorxiv.org/content/10.1...
(1/n)

Excited for a major milestone in our efforts to map enhancers and interpret variants in the human genome:

The E2G Portal! e2g.stanford.edu

This collates our predictions of enhancer-gene regulatory interactions across >1,600 cell types and tissues.

Uses cases 👇

1/

In the genomics community, we have focused pretty heavily on achieving state-of-the-art predictive performance.

While undoubtedly important, how we *use* these models after training is potentially even more important.

tangermeme v1.0.0 is out now. Hope you find it useful!

tangermeme: A toolkit for understanding cis-regulatory logic using deep learning models https://www.biorxiv.org/content/10.1101/2025.08.08.669296v1

Update of our protein outlier caller PROTRIDER. We now handle missing values, a widespread issue for mass spec where missing values are not a random -- and this improves outlier detection on non-missing data! Thumbs up to Daniela and George for the great work.
doi.org/10.1101/2025...

This year, the lab has a great representation at the #eshg2025: 3 talks, 2 posters, 1 spin-off stand ! 1/n

Our review "Predicting gene expression from DNA sequence using deep learning models" is finally out! 🤗

a fundamental challenge in my field is that staring at long-running jobs, waiting for them to finish, is not seen as productive

Join us for our next Kipoi Seminar with Laura Martens, Gagneur lab, TUM @lauradmartens.bsky.social @gagneurlab.bsky.social @tum.de
🐕scooby: Modeling multi-modal genomic profiles from DNA sequence at single-cell resolution
📅Wed May 7, 5:30pm CET
🧬https://kipoi.org/seminar/
🦋kipoizoo.bsky

Many of you enjoy our sequence-based model of single-cell RNA and ATAC data scooby... Don't miss Laura Marten's talk at the upcoming Kipoi seminar about it this Wed!
@lauradmartens.bsky.social @johahi.bsky.social @kipoizoo.bsky.social
Last preprint version:
www.biorxiv.org/content/10.1...

Programmatic design and editing of cis-regulatory elements The development of modern genome editing tools has enabled researchers to make such edits with high precision but has left unsolved the problem of designing these edits. As a solution, we propose Ledi...

Our preprint on designing and editing cis-regulatory elements using Ledidi is out! Ledidi turns *any* ML model (or set of models) into a designer of edits to DNA sequences that induce desired characteristics.

Preprint: www.biorxiv.org/content/10.1...
GitHub: github.com/jmschrei/led...

CREsted: modeling genomic and synthetic cell type-specific enhancers across tissues and species Sequence-based deep learning models have become the state of the art for the analysis of the genomic regulatory code. Particularly for transcriptional enhancers, deep learning models excel at decipher...

Very proud of two new preprints from the lab:
1) CREsted: to train sequence-to-function deep learning models on scATAC-seq atlases, and use them to decipher enhancer logic and design synthetic enhancers. This has been a wonderful lab-wide collaborative effort. www.biorxiv.org/content/10.1...

We released our preprint on the CREsted package. CREsted allows for complete modeling of cell type-specific enhancer codes from scATAC-seq data. We demonstrate CREsted’s robust functionality in various species and tissues, and in vivo validate our findings: www.biorxiv.org/content/10.1...

CREsted: modeling genomic and synthetic cell type-specific enhancers across tissues and species https://www.biorxiv.org/content/10.1101/2025.04.02.646812v1

In today's poster session #probgen25. To the pop gen folks, interesting observation: The influence of a nucleotide on reconstructing others, rather than its own reconstructability, is a better predictor of function. This metric makes DNA LMs beat conservation in several benchmarks.

and @pedrotomazdasilva.bsky.social will present tomorrow at #probgen25 poster 128 on dependency analysis of DNA language models. Come and see what functional relationships DNA LMs capture, from regulatory code to RNA structures. Preprint: doi.org/10.1101/2024...

Tomorrow Johannes Hingerl @johahi.bsky.social gives a talk on scooby at #probgen25. Enjoy learning in the legendary CSHL auditorium how to model RNA-seq and ATAC-seq profiles in individual cells from half a megabase of genomic sequence. Preprint:
doi.org/10.1101/2024...

Hello #probgen25! We have 3 contribs this year @lauradmartens.bsky.social starts today, poster 87, presenting scooby modeling scRNA-seq and sc-ATAC-seq profiles from DNA and applications. Shhh... don't tell it further... rumour says there are awesome cute scooby stickers to win ;-)

Join us for our next Kipoi Seminar with with Alexander Sasse
@lxsasse.bsky.social
@zmbh.uni-heidelberg.de

👉Advanced training strategies for genomic sequence-to-function models
📅 Wed March 5, 5:30pm CET
🧬 kipoi.org/seminar/
🦋 @kipoizoo.bsky.social

PROTRIDER: Protein abundance outlier detection from mass spectrometry-based proteomics data with a conditional autoencoder Motivation Detection of gene regulatory aberrations enhances our ability to interpret the impact of inherited and acquired genetic variation for rare disease diagnostics and tumor characterization. Wh...

Excited to share that PROTRIDER, our method to call outliers on mass spectrometry-based proteomics data, is out now!! #proteomics #massspectrometry #raredisease doi.org/10.1101/2025...

Just very happy to have our paper out today! A big thanks to all our co-authors, and to Nikolai and @steinaerts.bsky.social for the teamwork over the past years. If you are interested in using our models for cross-species enhancer studies, check out crested.readthedocs.io/en/stable/mo... 🙂

Can DNA sequence models predict mutations affecting human traits?

We introduce TraitGym, a curated benchmark of causal regulatory variants for 113 Mendelian & 83 complex traits, and evaluate functional genomics and DNA language models. Joint work w/ Gökcen Eraslan and @yun-s-song.bsky.social 🧵👇

Join us for our next Kipoi Seminar with with Pedro Tomaz da Silva @pedrotomazdasilva.bsky.social @gagneurlab.bsky.social @TU_Muenchen!
👉Nucleotide dependency analysis of DNA language models reveals genomic functional elements
📅Wed Feb 5, 5:30pm CET
🧬https://kipoi.org/seminar/
🦋kipoizoo.bsky