Machine Learning Talks on Campus

13.03.2026
4:00PM

Interpreting and designing regulatory DNA with deep learning
Assc. Prof. Dr. Peter Koo (Cold Spring Harbor Laboratory)

Genomic AI models have achieved strong performance in predicting functional genomics data, suggesting they capture key features underlying gene regulation. However, their black-box nature makes it challenging to determine whether they truly learn meaningful biological mechanisms. In this talk, I will present a suite of perturbation-based interpretation methods that leverage in silico mutations to identify the sequence features and combinatorial interactions that drive regulatory activity. These methods support interpretation across multiple scales—from individual nucleotides to entire cis-regulatory elements—revealing complex dependencies and generalizable cis-regulatory principles. In the second part of the talk, I will introduce DNA Discrete Diffusion (D3), a generative AI framework based on discrete diffusion for designing regulatory DNA sequences with tunable activity levels. D3 enables cell type-specific sequence generation, captures diverse regulatory patterns, and enhances the performance of predictive models when used to augment training data in low-data regimes. Together, these approaches demonstrate how predictive and generative AI can advance biological discovery by deepening our understanding of gene regulation and enabling the machine-guided design of regulatory sequences.

Dr. Alexander Sasse

Mathematikon 5.104 (INF 205), Heidelberg

Online: Zoom: https://eu02web.zoom-x.de/j/65716896158?pwd=5cQolLUW2AxUR6U3QU0h4YW8iOrYpg.1

 

15.06.2026 -19.06.2026

Workshop "Establishing a knowledge graph community in biomedical science" 

Many modern biomedical methods benefit from the availability of prior knowledge about, for example, genes, proteins, or diseases. Knowledge graphs, i.e., representations of prior knowledge in machine-readable graph form, have become the quasi-standard for storing, manipulating, and sharing biomedical prior knowledge.  

In this workshop, we will learn about knowledge representations, knowledge graphs, ontologies, and data structures, and put this knowledge to practical use with BioCypher (https://biocypher.org/). The workshop will also contain a module on information fusion, leveraging OntoWeaver (https://ontoweaver.readthedocs.io/en/latest/) to combine data from different sources and join this information into one single graph. Further, we will learn how to utilize generative AI to more generally develop our own data collection, transformation, and combination workflows. 

Registration is free and open until May 15th, 2026: https://biocypher.org/community/2026-workshop/ 

Inga Ulusoy, Scientific Software Center

INF 205, Mathematikon, 5/104 (Conference Room)