scRNA-seq Differential Expression

Introduction

This dataset consists of gene expression data of human pancreatic islet cells of both healthy and diabetic donors. The aim of the study is to discover differences in gene expression between healthy and diabetic cells.

Dataset description

  • Organism: Homo sapiens

  • Instrument: Illumina HiSeq 2500

  • Library construction: SMARTer v1

  • Layout: Single-end 75 pb

  • Number of cells: 1,068

Publication

Lawlor N, George J, Bolisetty M, et al. Single-cell transcriptomes identify human islet cell signatures and reveal cell-type-specific expression changes in type 2 diabetes. Genome Research. 2017 Feb;27(2):208-222. DOI: 10.1101/gr.212720.116. PMID: 27864352; PMCID: PMC5287227.

Bioinformatic Analysis

1.- scRNA-Seq Clustering

Application

Single-cell RNAseq Clustering

Input

Parameters

  • Input Type: Count Table File

  • Count Table File: counts_islets.txt

  • Column Separator: TAB

  • NA Values: Assume Zero Values

  • Minimum Cells: 4

  • Set Minimum Counts Cutoff: false

  • Set Maximum Counts Cutoff: true

  • Maximum Counts: 5000000

  • Set Minimum Features Cutoff: false

  • Set Maximum Features Cutoff: false

  • Filter by % of Mitochondrial Genes: false

  • Multi Sample Analysis: true

  • Experimental Design File: exp_design_islets.txt

  • Condition: disease

  • N. Dimensions for Integration: 20

  • K Anchor: 5

  • K Filter: 200

  • K Score: 30

  • K Weight: 100

  • Normalize Data: true

  • Normalization Method: Regularized Negative Binomial Regression

  • High Variable Features: 3000

  • Scale Data: false

  • Center Data: true

  • Regress Out Mitochondrial Genes: false

  • Regress Out Cell Cycle Genes: false

  • Principal Components: 50

  • Define Dimensions by: Manual

  • Number of Dimensions: 12

  • k-value: 20

  • Resolution: 0.8

  • Point's Minimum Distance: 0.3

  • Point's Spread: 1.0

Execution Time

~ 10 min

Output

2.- scRNA-Seq Differential Expression

Application

Input

Parameters

  • Filtering Mode: Counts Per Million

  • CPM Filter: 0.0

  • Cells Reaching CPM Filter: 1

  • Normalization Method: TMM with Zero Pairing

  • Biological Replicates: individual

  • Diffexp Option: Simple Design

  • Primary Factor: clusters

  • Primary Contrast Conditions: cluster_1,cluster_9,cluster_2,cluster_3,cluster_6,cluster_5,cluster_4,cluster_8,cluster_7

  • Primary Reference Conditions: cluster_1,cluster_9,cluster_2,cluster_3,cluster_6,cluster_5,cluster_4,cluster_8,cluster_7

  • Blocking Factor: disease

  • Test Contrasts Separately: true

  • Select a Statistical Test: GLM (Likelihood Ratio Test)

  • Robust: true

Execution Time

~ 10 min

Output