Related Experiment Video
Updated: Jun 14, 2026
09:33
Diffusion Tensor Magnetic Resonance Imaging in the Analysis of Neurodegenerative Diseases
Published on: July 28, 2013
28.4K
Adaptive digital tissue deconvolution
Franziska Görtler1,2, Malte Mensching-Buhr3, Ørjan Skaar4
1Computational Biology Unit, Department of Biological Sciences, University of Bergen, N-5008 Bergen, Norway.
Bioinformatics (Oxford, England)
|June 28, 2024
Summary
Adaptive Digital Tissue Deconvolution (ADTD) improves cell type estimation from transcriptomics data by accounting for unknown cell contributions and adapting reference profiles. This method enhances accuracy and provides new insights into cell-specific differences in diseases like breast cancer.
Area of Science:
- Computational biology
- Genomics
- Bioinformatics
Background:
- Estimating cellular composition from transcriptomics data is crucial for biological insights.
- Existing computational methods, including machine learning, struggle with unknown cell types and fixed reference profiles.
- Cellular phenotypes adapt to their environment, challenging the accuracy of generic cell archetype references.
Purpose of the Study:
- To develop a novel computational approach, Adaptive Digital Tissue Deconvolution (ADTD), for accurate cell type proportion estimation.
- To address limitations of existing methods by handling unknown cellular contributions and adapting reference profiles.
- To uncover cell-type specific gene regulation and molecular differences from bulk transcriptomics data.
Main Methods:
- Adaptive Digital Tissue Deconvolution (ADTD) algorithm.
- Utilizes machine learning to estimate cell proportions and unknown background contributions.
- Adapts prototypic reference profiles to the cellular molecular environment.
Main Results:
- ADTD accurately estimates cellular compositions, outperforming existing methods in simulations.
- The method successfully infers unknown and hidden cellular contributions.
- Application to breast cancer data revealed cell-type specific molecular differences across subtypes.
Conclusions:
- ADTD offers a robust solution for deconvolution of transcriptomics data, handling complex biological scenarios.
- The adaptive nature of ADTD improves the resolution of cell-type specific gene expression.
- This approach provides valuable insights for understanding disease mechanisms and subtypes.
