Research Field
Computational Biology
AI models uncover the regulatory grammar of nucleic acids across genomes, cells, and populations.
Description
Advances in sequencing technologies now capture diverse layers of nucleic acid biology, from genome sequence and epigenetic regulation to gene expression, RNA processing, and protein–RNA interactions. This research area seeks to integrate these molecular modalities to uncover the principles that govern gene regulation, genetic variation, and disease.
A central goal is to develop advanced AI models capable of learning meaningful representations of nucleic acid sequences and their regulatory contexts across multiple biological scales. The research combines large-scale genomic, transcriptomic, epigenomic, and single-cell datasets to identify regulatory mechanisms, interpret disease-associated genetic variation, and link molecular phenotypes to cellular function. Particular emphasis is placed on integrating multimodal data and developing foundation models that capture the underlying grammar of nucleic acid regulation. By transforming complex biological datasets into predictive frameworks, the project aims to reveal previously inaccessible regulatory principles and accelerate the discovery of disease mechanisms and therapeutic targets.
Area C1
Publications
SAM68 is a multifunctional post-transcriptional regulator of cardiomyocyte differentiation.
Nucleic Acids Res. · 2026
Show abstract
RNA-binding proteins (RBPs) of the STAR family play key roles in mammalian development, yet their contributions to lineage specification remain incompletely understood. Here, using CRISPR-Cas9 knockout models combined with multi-omics approaches, we investigate the functions of two STAR proteins, SAM68 and QKI, in mouse embryonic stem cells (mESCs). Both RBPs support mESC proliferation, self-renewal, and efficient differentiation into cardiomyocytes. Although SAM68 and QKI belong to the same protein family, they control largely distinct regulatory programs during differentiation. We uncover an unexpected role for SAM68 in cardiomyocyte specification through multiple post-transcriptional mechanisms. SAM68 modulates alternative splicing and promotes the biogenesis of a subset of cardiac-enriched circular RNAs, through binding to intronic regions flanking back-splice junctions and potentially through association with NF90/110. In addition, SAM68 binds untranslated regions of key differentiation-related transcripts, including Gata4 mRNA, and functions in ribonucleoprotein complexes to regulate their translation. Together, these findings identify SAM68 as a multifunctional regulator coordinating multiple layers of RNA metabolism-including splicing, circRNA biogenesis, and translation-during cardiomyocyte differentiation and provide insight into how STAR proteins shape post-transcriptional gene regulatory networks during early development.
RNA motifs, RNA structure, and motif context analyzed by RNAanalyzer3.
Nucleic Acids Res. · 2026
Show abstract
RNAanalyzer3 ("RNA analyzer cubic"; https://rnaanalyzer.bioapps.biozentrum.uni-wuerzburg.de) substitutes the frequently consulted current RNAanalyzer webserver (https://rnaanalyzer-old.bioapps.biozentrum.uni-wuerzburg.de). RNAanalyzer3 is free/open via the secure HTTPS protocol, with example data, help and tutorial, web-link to results, and rich data output. We combine a general detailed structure analysis with motif analyses. It accepts either a single plain-text nucleotide sequence or batch submission in FASTA format, which can be pasted or uploaded as a FASTA file. Our tool (i) has up-to-date software and operating systems, (ii) combines diverse RNA motif analyses with RNA structure prediction, (iii) puts found motifs into structural context, and (iv) offers dedicated tools for probing RNA-protein binding interactions. RNAanalyzer3 links motif searches to Rfam and miRNA search to miRbase. It focuses on structural features first, looks for stem-loops, hairpins, and specific enrichment regions such as stem-GG pairs, plus AU-rich regions with their locations for easier identification, while providing structural context and interactive RNA structure visualization. A tabulated overview shows all RNA features including structure details, coding potential, untranslated regions (UTRs, including Shine-Dalgarno sequences, Kozak sequences, and polyadenylation signals), transfer RNA (tRNA), microRNA (miRNA), long noncoding RNA (lncRNA), trans-splicing motifs, iron response elements (IRE), riboswitches, small nuclear ribonucleoprotein (snRNP) motifs, and spliceosomal Sm-sites.