Main functions embedded in FlowCT and their application
| . | Embedded functions . | Utility . |
|---|---|---|
| Quality control | ||
| flowAI | flow_auto_qc | Removes low-quality events by evaluating flow rate, signal acquisition, and dynamic range. |
| Marker normalization | ||
| flowStats | gaussNorm & warpSet | Normalizes flow cytometry data sets by aligning high-density regions (ie, landmarks or peaks) for each channel. |
| Seurat | SelectIntegrationFeatures & IntegrateData | Identifies anchors between pairs of data sets and uses them to remove confounding factors. |
| harmony | HarmonyMatrix | Corrects batch effects through a maximum diversity algorithm (ie, soft k-means) and a mixture model–based linear correction. |
| Automatic clustering | ||
| FlowSOM | BuildSOM & ConsensusClusterPlus | Creates clusters from flow cytometry data sets based on self-organizing map (SOM) and minimal spanning trees (MSTs). |
| PARC | PARC | Identifies single-cell clusters through a combination of graph-based clustering and pruning, coupled with the Leiden community-detection algorithm. |
| Rphenograph | Rphenograph | Clusters single cells by using the Louvain method based on a previous phenotypically defined graph. |
| Seurat | FindNeighbors & FindClusters | Finds single-cell communities based on k-nearest neighbor (KNN) graphs and clustering via Louvain or smart local moving (SLM) algorithms. |
| Dimensionality reduction | ||
| Rtsne | Rtsne | Calculates t-distributed stochastic neighbor embedding (t-SNE). |
| uwot | tumap | Calculates uniform manifold approximation and projection (UMAP). |
| densvis | densmap & densne | Produces lower-dimensional embeddings (t-SNE- and UMAP-based) preserving the density of cells. |
| Machine learning | ||
| biosigner | biosign | Features selection by running partial least squares-discriminant analysis (PLS-DA), random forest, and support vector machine (SVM) simultaneously (all methods as binary classifiers). |
| randomForestSRC | rfsrc | Selects immune populations based on random forest building and incorporates survival information. |
| SurvBoost | boosting_core | Detects more relevant populations through gradient boosting algorithm and includes survival data. |
| . | Embedded functions . | Utility . |
|---|---|---|
| Quality control | ||
| flowAI | flow_auto_qc | Removes low-quality events by evaluating flow rate, signal acquisition, and dynamic range. |
| Marker normalization | ||
| flowStats | gaussNorm & warpSet | Normalizes flow cytometry data sets by aligning high-density regions (ie, landmarks or peaks) for each channel. |
| Seurat | SelectIntegrationFeatures & IntegrateData | Identifies anchors between pairs of data sets and uses them to remove confounding factors. |
| harmony | HarmonyMatrix | Corrects batch effects through a maximum diversity algorithm (ie, soft k-means) and a mixture model–based linear correction. |
| Automatic clustering | ||
| FlowSOM | BuildSOM & ConsensusClusterPlus | Creates clusters from flow cytometry data sets based on self-organizing map (SOM) and minimal spanning trees (MSTs). |
| PARC | PARC | Identifies single-cell clusters through a combination of graph-based clustering and pruning, coupled with the Leiden community-detection algorithm. |
| Rphenograph | Rphenograph | Clusters single cells by using the Louvain method based on a previous phenotypically defined graph. |
| Seurat | FindNeighbors & FindClusters | Finds single-cell communities based on k-nearest neighbor (KNN) graphs and clustering via Louvain or smart local moving (SLM) algorithms. |
| Dimensionality reduction | ||
| Rtsne | Rtsne | Calculates t-distributed stochastic neighbor embedding (t-SNE). |
| uwot | tumap | Calculates uniform manifold approximation and projection (UMAP). |
| densvis | densmap & densne | Produces lower-dimensional embeddings (t-SNE- and UMAP-based) preserving the density of cells. |
| Machine learning | ||
| biosigner | biosign | Features selection by running partial least squares-discriminant analysis (PLS-DA), random forest, and support vector machine (SVM) simultaneously (all methods as binary classifiers). |
| randomForestSRC | rfsrc | Selects immune populations based on random forest building and incorporates survival information. |
| SurvBoost | boosting_core | Detects more relevant populations through gradient boosting algorithm and includes survival data. |