Data

In this menu you can select a dataset that you want to analyze within the UMAP menu ( ). Once you have chosen a dataset, make sure you click "Load" in order to make it available in the UMAP menu. After you loaded a dataset, a brief description of the dataset with regard to its experimental details as well as a preview of the UMAP embedding is shown. You are now set up to explorer the selected dataset with GutTubeR.

UMAP

In this menu you can create UMAPs of the selected dataset in a dynamic fashion. First, select how many plots you want to create by activating up to four plots with the switches at the top. After having done so, you can select the meta annotations for the corresponding UMAPs from a drop-down menu. Note, if you want to annotate gene expression, just type and search for your gene of interest in the selection field. You can also filter grouping annotations by selecting multiple groups from the filter selection menu. After having chosen a annotation click the Add button ( ). Proceed like this for all activated plots. Once you set all the annotations you want to plot, click the Update button ( ) and your UMAPs will be rendered. Note, that when clicking the Add button ( ) the corresponding plot will be created with the current settings and it will overwrite existing annotations for that plot. Adjusting settings ( ) later on will just affect newly added plots. This allows you to select for example different color scales for gene expression on the same UMAP or load and compare other datasets to already created UMAPs from different datasets. You can download the plots you created by clicking the Download button ( ) with the set download parameters.

Performance

Some of the available datasets are rather huge which leads to increased plotting time. When you encounter such behavior, you can activate the performance mode ( ). When activated the selected dataset is sampled before creating the UMAP, which reduces overplotting and plotting time. The applied sampling rate can be adjusted in the settings ( ). By default the performance mode is activated ( ) when the selected dataset exceeds 150,000 cells otherwise it is deactivated ( ).

Statistics

The statistics pop-up window allows you to plot underlying distributions of your created UMAPS. You can either accompany a selected UMAP with a distribution for its selected annotation or compare two UMAPs and their corresponding distributions. Note, that when you select a UMAPs with a discrete mapping and a gene expression mapping, a combined violine plot will be created showing the gene expression grouped by the selected discrete meta annotation. In many cases, it makes sense to filter the gene expression values for not expressed genes. You can toggle this with the clip gene expression histogram option. You can download the displayed distribution plots by clicking the Download button ( ). Here the same download parameters apply as for the UMAP section ( )

Settings

In the settings menu you can specify options for plotting and downloading. The chosen settings apply for the main UMAP plotting functionality ( ) as well as for the statistics menu ( ). By now you can choose a color scale for continous colormapping (blues and viridis color palettes) of gene expression, set the point size and set the sampling rate (0.1 - 1) which is applied when the performance mode ( ) is activated. Apart from this you can specify aspect ratio, unit, resolution and format parameters for downloading ( ) the created UMAPs and distribution plots as an image file.

---

Browser

This sections allows you to browse the results of CellPhoneDB analysis for signalling between mesenchyme and epithelial across all organs available from our datasets with an interactive interaction heatmap. In order to start select a dataset and the signalling direction and click the Update button ( ). Please be patient when loading the interaction heatmap for the All Organs dataset. Once the heatmap is loaded, you can start hovering over it in order to highlight specific ligand-receptor pairs and their related pathway/celltype/organ information. When you found a interesting pair, click on it in the heatmap and click Add ( ) in order to save this interaction and investigate it further in the Inspector section ( ). Note that you can navigate to specific areas of the heatmap by drawing a rectangular. This will zoom into the selected area. Double-clicking will show the entire heatmap again. It is also possible to filter the interaction pairs for genes of your interest in the Settings ( ) as well as setting the center of the colorscaling to either rowwise or columnwise. Changed settings will apply as soon as you click Update ( ) again. You can download the displayed heatmap by clicking on download button ( ) in the upper right corner of the interactive heatmap.

Inspector

This section consists of two subsections: The Editor ( ) that allows you to view and edit all selected interactions from the Browser section ( ) as well as add receptor-ligand pairs manually. And Inspect ( ) which enables you to map ligand-receptor pairs by bivariate color mapping to UMAPs of either all organs or specific ones. The plotting interface has the same functionalities as already described for the Atlas section ( ). Meaning that you can plot up to four different UMAP embeddings of different datasets with annotations for gene expression for either ligand or receptor, age, organ, tissue and cell type. In addition it is possible to select annotations for ligand-receptor expression which utilizes bivariate coloring and mapping. Within the statistics menu ( ) you can view the underlying distributions of UMAPs created with the dynamic plotting interface.

In the Statistics pop-up window ( ) you can accompany the UMAP of selected ligand-receptor interactions or other selected annotations with their corresponding distributions grouped by either organ, tissue or cell type. When a grouping is selected, the corresponding annotated UMAP is plotted additionally. Note, that having to many groups can create faceted histograms that are not feasible. When you encounter this, please filter the displayed groups from the selection menu. The filtering also applies to the added UMAP with the grouping annotation. In general, the same setting options ( ) apply for the Inspector interface ( ) as for the Atlas section ( ) with regards to setting colorscales, point size, performance mode, sampling rate and download parameters.

---