Cryo-electron tomography derived cell population analysis through novel high-throughput machine learning approaches
Active Project Grant Awarded Last action
- Award ID
- R35GM158094
- Recipient
-
Carnegie-Mellon University UEI U3NKNFLNQ613
- Awarding agency
-
National Institute of General Medical Sciences
National Institutes of Health, Department of Health and Human Services
- Obligated
- $742,854
Award facts
- Assistance listing
- 93.859 Biomedical Research and Research Training
- Completion date 29% complete
- Place of performance
- Pittsburgh, PA 15213, USA
Award details come from USAspending.gov.
Description
- Abstract
- Cryo-electron tomography (cryo-ET) has emerged as a major tool for in situ biology which enabled the system- atic 3D visualization of subcellular objects and their spatial organization. Cryo-ET enables recovering entire subcellular morphologies and organizations across different cell populations, bringing significant potential for downstream biomedical tasks of phenotype-genotype analysis, disease diagnosis, and drug design. A major challenge in utilizing the potential of cryo-ET for such downstream tasks is the lack of methods to systematically study the cell-population-wide differences in subcellular object morphology and their spatial organization. The goal of this program is to develop novel high-throughput machine-learning approaches to study cell-population- wide differences in subcellular object morphology, conformations, and organization. We set out to solve multiple computational problems toward this goal. The current deep learning-based cellular cryo-ET tomogram analysis approaches have limited data-efficiency and apply to one or a few tomograms having similar data domains. Consequently, they are unsuitable for large-scale analysis of tomograms of diverse data domains that can be used to image a cell population. In our first module, we will develop domain generalization methods to IM- prove the data efficiency of existing deep learning-based cryo-ET image analysis methods. Our next module will develop high-throughput methods for coarse-to-fine segmentation of multi-scale subcellular objects in cryo- ET tomograms with unsupervised and weakly-supervised learning. In the following module, we will develop novel representation learning methods for statistical shape and morphology analysis of the subcellular objects. Finally, we will develop interpretable methods to study cell population-wide shape and spatial traits of multi- scale subcellular objects with multiple biological CO-variates. In addition to the novel algorithms, we will create user-friendly software with a graphical user interface (GUI) incorporating the algorithms that will ensure the biologists can leverage cryo-ET for cell-population analysis.
Transactions
| Action | Obligated | Date |
|---|---|---|
| Amendment | $371,427 | |
| Original award | $371,427 |
Quick answers
Who received award R35GM158094?
R35GM158094 was awarded to Carnegie-Mellon University by National Institute of General Medical Sciences (National Institutes of Health) on June 10, 2025.
How much is award R35GM158094 worth?
$742,854 has been obligated on R35GM158094.
On GovTribe
See this award's full record
- Every transaction and subaward
- The recipient's other awards and teaming partners
- Similar grant awards
- Ask GovTribe AI about this award