2022 Fiscal Year Research-status Report
Inferring cells differentiation processes from single-cell Multiome ATACseq+RNAseq data.
| Project/Area Number |
22K21301
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2022-08-31 – 2024-03-31
|
| Keywords | Bioinformatics / Cells-reprogramming / Multiomics / RNA-velocity |
|---|
| Outline of Annual Research Achievements |
I confirmed the significance of this project in terms of a recent method published in a high-impact factor journal.
I set up the computational environment and worked on the estimation of gene expression levels of cells within clusters.
I tested different linear methods with and without regularization terms. I conclude that estimation with regularization, especially the ridge regressor, was more appropriate for single-cell RNA-seq data as they reduce the impact of the high sparsity of the data.
I set up the fuzzy logic interpolation. I performed different tests on the interpolation of the estimated cluster-specific gene dynamics with satisfactory results. With these tests, I was able to perform in-silico simulations of a cell differentiating from one cluster to another.
|
| Current Status of Research Progress |
Current Status of Research Progress
3: Progress in research has been slightly delayed.
Reason
Recently, a method related to my proposed research was published in a high-impact journal. Then, I spent the first couple of months analyzing the results and pipeline of the newly published method to confirm the significance of this project's expected results. In the end, I confirmed that my proposed pipeline could offer significant results but lost considerable time. However, seeing a similar technique published in a high-impact factor was motivating as it made me realize the importance of this project.
|
| Strategy for Future Research Activity |
I want to compare the fuzzy logic using pseudo-time as the dynamic link between clusters. I expect that this change will help to improve the interpolation process.
The next step will be to incorporate the data from ATAC-seq data into the dynamical model. For this, I will first integrate the cells from ATAC and RNA-seq data to set a common link between cells from different data types, in this case, the pseudo-time. Then, I will apply the same estimation and fuzzy logic procedures as with the RNA-seq data.
Once I obtain the complete model, RNA + ATAC-seq, I will be able to perform an in-silico simulation of cell differentiating.
|
| Causes of Carryover |
I didn't have the opportunity to join another local meeting during the past year. The remaining budget will be used to attend two international meetings on this fiscal year, which are usually more expensive.
|
