Chapter 7 Structure of CytoTree
The CYT object is according to the R S4 object. The raw.data which consists of the imput matrix data information is required to build the CYT object. When running the CytoTree::createCYT function, log.data in the CYT object is the log-transformed matrix and is filtered according to the markers included in raw.data. The log.data is essential for the next workflow. The slots som and cluster contain clustering information and parameter of clustering module, whereas pca.sdev, pca.value, pca.score, tsne.value, dm, umap.value store the running parameters and results of dimensionality reduction module. The network built by MST is stored in network slots. The root cells and leaf cells are stored in the slots root.cells and leaf.cells and are essential for pseudotime estimation and trajectory inference. The slots knn, knn.index and knn.distance contain the KNN information calculated by the R BiocNeighbors package. And the slots walk and diff.tree contains the intermediate calculation results in the pseudotime and trajectory analysis. The meta-information for cells and clusters are stored in the slots plot.meta and tree.meta, which can be fetched by CytoTree::fetchPlotMeta and CytoTree::fetchClustMeta, respectively. All slots are constrained based on the standardized data type in R.
| Function | Slot | Description | Type |
|---|---|---|---|
| Basic | raw.data | Raw signal data captured from FCS | matrix |
| Basic | log.data | Data enrolled in the computational modules | matrix |
| Basic | meta.data | Meta data information of the experiment, and columns of “stage” and “cell” are required | data.frame |
| Basic | markers | Markers used in the calculation of PCA, tSNE, diffusion map and UMAP. | vector |
| Basic | markers.idx | Index of markers | vector |
| Basic | cell.name | Cell names after downsampling | vector |
| Clustering | som | Store som network information calculated using FlowSOM | list |
| Clustering | cluster | Store clustering information after processing cell clusters | data.frame |
| Dimensionality Reduction | pca.sdev | Storing PCA information | vector |
| Dimensionality Reduction | pca.value | Storing PCA information | matrix |
| Dimensionality Reduction | pca.scores | Storing PCA information | matrix |
| Dimensionality Reduction | tsne.value | Storing tSNE information | matrix |
| Dimensionality Reduction | dm | Storing diffusion maps information | S4 object |
| Dimensionality Reduction | umap.value | Storing UMAP information | matrix |
| Pseudotime | root.cells | Root cells | vector |
| Pseudotime | leaf.cells | Leaf cells | vector |
| Pseudotime | network | Minimum spanning tree information | list |
| Pseudotime | knn | Numbers of nearest neighbors | numeric |
| Pseudotime | knn.index | Matrix corresponds to a point in log.data and contains the row indices in log.data | matrix |
| Pseudotime | knn.distance | Contains the distance of nearest neighbors of each cell | matrix |
| Intermediate state | walk | Walk parameters between leaf cells and root cells | list |
| Intermediate state | diff.tree | Storing walking information | list |
| Meta information | plot.meta | Meta information for single cells in visualization | data.frame |
| Meta information | tree.meta | Meta information for clusters in visualization | list |