# Perform Exploratory analysis

* [Use Principle Components Analysis (PCA) to reduce dimensions](#use-principle-components-analysis-pca-to-reduce-dimensions)
* [Classify cells based on a marker for expression](#classify-cells-based-on-a-marker-for-expression)

## Use Principle Components Analysis (PCA) to reduce dimensions

* Click the **Normalized counts** data node
* Expand the **Exploratory analysis** section of the task menu
* Click **PCA**

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-33da68fc5b8f48d5a29dab3c404c906158f2b17b%2Fimage2023-7-21_17-26-37.png?alt=media)

In this tutorial we will modify the PCA task parameters, to not split by sample, to keep the cells from both samples on the PCA output.

* Uncheck (de-select) the **Split by sample** checkbox under *Grouping*
* Click **Finish**

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-bdd4b960b9afdacbaebc72ac0c1d2d11cc412f1f%2Fimage2023-7-21_17-29-18.png?alt=media)

* Double-click the circular **PCA** node to view the results

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-2cc98588ec0dcd22435169aaa096f51f0cfa0fb4%2Fimage2023-7-21_17-36-25.png?alt=media)

From this PCA node, further exploratory tasks can be performed (e.g. t-SNE, UMAP, and Graph-based clustering).

## Classify cells based on a marker for expression

* Choose **Style** under *Configure*
* **Color by** and search for *fasn* by typing the name
* Select *FASN* from the drop-down

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-a9b8e0bf896b8844f28363687eba8d08b0a1743c%2Fimage2023-7-21_17-50-30.png?alt=media)

The colors can be customized by selecting the **color palette** then using the color drop-downs as shown below.

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-48f52bb7ebc554e10652b59ebe6e84a395f3deea%2Fimage2023-7-21_18-3-30.png?alt=media)

Ensure the colors are distinguishable such as in the image above using a blue and green scale for **Maximum** and **Minimum**, respectively.

* Click **FASN** in the legend to make it draggable (pale green background) and continue to drag and drop FASN to **Add criteria** within the **Select & Filter** *Tool*
* Hover over the slider to see the distribution of FASN expression

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-6238f0b9c22a7caf0db0cc280beff4583941dd4a%2Fimage2023-7-21_18-24-33.png?alt=media)

Multiple gene thresholds can be used in this type of classification by performing this step with multiple markers.

* Drag the slider to select the population of cells expressing high FASN (the cutoff here is 10 or the middle of the distribution).

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-9f88a7b0fd1a4650e9e379e65a3b17e9de153e10%2Fimage2023-7-21_18-28-32.png?alt=media)

* Click **Classify** under *Tools*
* Click **Classify selection**

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-2fc64845dba870180ad1b2aa30d5159b0c41a227%2Fimage2023-7-21_18-31-4.png?alt=media)

* Give the classification a name "FASN high"

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-0934218b9fafdaab7ff952d468355bb02c3e98cb%2Fimage2023-7-21_18-36-49.png?alt=media)

* Under the Select & Filter tool, choose **Filter** to **exclude** the selected cells

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-831e6eaa9b4638cfcdc7b36fa802962b673d983a%2Fimage2023-7-21_18-39-32.png?alt=media)

Exit all Tools and Configure options

* Click the "X" in the right corner
* Use the **rectangle** selection mode on the PCA to select all of the points on the image

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-ccddac6a8edb574e8aec8cfb16e20cc3dc84beac%2Fimage2023-7-21_18-43-15.png?alt=media)

This results in 147538 cells selected.

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-8da1b718124d362b1341115e7aed2a56196d859e%2Fimage2023-7-24_15-58-4.png?alt=media)

* Open **Classify**
* Click **Classify selection** and name this population of cells "FASN low"
* Click **Apply classifications** and give the classification a name "FASN expression"

![](https://1384254481-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJVEESmJAPppJ3ijFq5aR%2Fuploads%2Fgit-blob-48f006da0fdcfc0262b1c9ecbdebadb8f6f6c393%2Fimage2023-7-24_16-1-1.png?alt=media)

Now we will be able to use this classification in downstream applications (e.g. differential analysis).

## Additional Assistance

If you need additional assistance, please visit [our support page](http://www.partek.com/support) to submit a help ticket or find phone numbers for regional support.