Blue:RNA-Seq Analyses:C1 hNuc 1000 Dataset: Difference between revisions

From ZhangLabWiki
Jump to navigation Jump to search
>B1lake
>B1lake
Line 33: Line 33:


*Top 27 genes based on [http://genome-tech.ucsd.edu/LabNotes/index.php/Rui:LabNotes/SingleCell/2014-7-4#First_round_of_sample_clusters_based_on_overdispersed_genes Rui's] first round of sample clustering:
*Top 27 genes based on [http://genome-tech.ucsd.edu/LabNotes/index.php/Rui:LabNotes/SingleCell/2014-7-4#First_round_of_sample_clusters_based_on_overdispersed_genes Rui's] first round of sample clustering:
**Updating gene information and associate references


[[File:Gene_Clustering_Analyses_Round_1_Top_27.xlsx]]
[[File:Gene_Clustering_Analyses_Round_1_Top_27.xlsx]]

Revision as of 00:58, 12 July 2014

Overview

  • C1 Run Summary:

File:20chips Samplelist.png

  • Primary Mapping/Analysis done by Rui using our pipeline
  • More in depth analysis by Rui

Average cDNA Yields Across Runs

File:20 C1 Runs cDNA Yields 2.png


  • Circled experiments represent the same batch of nuclei. There is little effect of freezing the nuclei in 10%DMSO or running the nuclei on a small chip rather than the medium chip. Global levels of cDNA were reduced with adding a nuclei stain/wash step to the loading protocol using the DNA Seq Cell Wash buffer. Also, addition of PolydIdC did globally increase cDNA levels even for the 0_Nuc conditions.
  • Examination of basic mapping stats shows that the small chip run 19 was highly variable (likely due to capture of small low RNA containing nuclei), while addition of PolydIdC seemed to stabilize mapping rates. Also, even though the cDNA yields were very high for the 0_Nuc with PolydIdC, most of the reads still mapped to ERCC.

Comparison of C1 Run Variations on the Same Nuclei Batch

  • Using Rui's mapped TPM data, performed analysis using Fluidigmn's Singular software:
    • Identified outliers based on log2 mapping rates and trimmed genes with TPM less than 2 (LOD1)
    • Ran Auto Analysis for top 100 differentially expressed genes
      • Performed pairwise comparison of this dataset:
      • Performed PCA analysis for this dataset:


File:0227 Nuc Pairwise 1.png


File:0227 Nuc PCA 1.png

  • Can see that none of the conditions significantly affected the gene expression profiles (using top 100 differentially expressed genes) of the captured nuclei and these nuclei retained similar PCA distributions across conditions

Analyses of Gene Clusters

  • Top 27 genes based on Rui's first round of sample clustering:
    • Updating gene information and associate references

File:Gene Clustering Analyses Round 1 Top 27.xlsx