Disputation: Ping-Han Hsieh

Doctoral candidate Ping-Han Hsieh at the Department of Informatics, Faculty of Mathematics and Natural Sciences, is defending the thesis Developing Computational Methods to Improve Gene Regulatory Network Inference for the degree of Philosophiae Doctor.

Bildet kan inneholde: briller, hår, smil, briller, hake.
Foto: Privat

The PhD defence will be partially digital, in Kristen Nygaards sal (5370), Ole-Johan Dahls hus and streamed directly using Zoom. The host of the session will moderate the technicalities while the chair of the defence will moderate the disputation.

Ex auditorio questions: the chair of the defence will invite the attending audience at Kristen Nygaards sal to ask ex auditorio questions. 

Trial lecture

" Challenge of the decade: Can gene function be derived from multiomics analysis?”

Time and place: October 9,  2024 11:15 AM. Kristen Nygaards sal (5370), Ole Johan Dahls hus.

Main research findings

Transcriptional regulation is a biological mechanism that requires the coordination of diverse regulatory elements and transcription factors to mediate gene transcription. Understanding these processes reveals molecular dynamics governing cellular responses to environmental conditions, diseases, and developmental stages. While many computational methods model these regulatory mechanisms as networks, challenges persist in their reproducibility and applicability.

This study addresses key challenges in regulatory network inference. We introduce SNAIL, an improved data normalization method that preserves the biological variability in the expression profile while reducing false-positive associations in the resultant networks. We have also implemented lionessR as a Bioconductor R package, enhancing the accessibility of the LIONESS algorithm for the broader scientific community. This package enables researchers to infer sample-specific regulatory networks from population-level data with ease. Lastly, we develop CAVACHON, a probabilistic learning framework based on hierarchical variational autoencoders that integrates multi-modal molecular data and prior biological knowledge for a more comprehensive understanding of regulatory dynamics.

These contributions aim to enhance the reproducibility, applicability, and biological relevance of regulatory network inference methods, ultimately deepening our understanding of cellular processes and biological systems.

Adjudication committee

  • Professor Esti Yeger-Lotem, Ben-Gurion University of the Negev, Israel
  • Associate Professor Johan Henriksson, Umeå University, Sweden
  • Professor II Rolf Inge Skotheim, University of Oslo, Norway

Supervisors

  • Head of the Kuijjer Research group, PhD, Marieke Kuijjer, Department of Informatics, University of Oslo, Norway

  • Professor Geir Kjetil Ferkingstad Sandve, Department of Informatics, University of Oslo, Norway

Chair of defence

Associate Professor Ellen Munthe-Kaas

Contact information at Department: Mozhdeh Sheibani Harat 

Publisert 24. sep. 2024 09:23 - Sist endret 8. okt. 2024 08:40