Scanpy Practice - PBMC Preprocessing & Clustering

---

Background:

Coming from 5 years of research experience, I have a growing interest for the field of Bioinformatics. As such I have been exploring personal projects & practices to apply computational methods to biological data.

Scanpy is a very powerful tool that I have come across. Within the toolkit's API & resources, Scanpy provides a tutorial section, which I work through here to not only explore my personal interests, but also apply computational methods to real genomics data. I expand on this tutorial by applying knowledge of the Scanpy toolkit via annotations as well as altering certain aspects of the analysis.

---

Context:

Scanpy is a Python-based toolkit used for analyzing single-cell gene expression data, and can be used for large-scale data analysis which includes preprocessing, visualizations, clustering, and differential expression testing.

FEATURES:

• Preprocessing: Doublet detection to identify and remove cells that result in misclassifications.
• Visualizations: Plotting functions
• Clustering: Methods for identifying clusters in the data
• Trajectory Inference: Methods for inferring trajectories in the data
• Differential Expression Testing: Methods for identifying differentially expressed genes
• Simulations: Methods for simulating gene regulatory networks

---

Walk-Through:

Within this tutorial, there is a focus on:

1. Preprocessing

2. Principal Component Analysis

3. Clustering

4. Marker Gene Identifications