Skip to content

In_vitro_analysis_steps.md

Latest commit

 

History

History
60 lines (55 loc) · 4.41 KB

In_vitro_analysis_steps.md

File metadata and controls

60 lines (55 loc) · 4.41 KB

In vitro screen data analysis

0. Data compilation

  • 0.1 Identify Amtrine+ / Amtrine+GFP+ columns (transduced population) in original spreadsheet
  • 0.2 Extract GeoMean / Percentage data for each marker for Amtrine+ / Amtrine+GFP+
  • 0.3 Write output files for each individual marker including: plate and well position and Flow readout (MFI or percentage)
  • 0.4 Match shRNA names to position

Script: 0_dataPrep.py
Source files: Screen_markers
Output files:
1_0_Raw
1_1_shRNAmatched

1. Data normalization

  • 1.1 Calculate percentile and Z-score of each sample
    • 1.1.1 Approach 1: Calculate normal distribution percentile and Z-score based on total population
    • 1.1.2 Approach 2: Calculate normal distribution percentile and Z-score based on control plate
  • 1.2 Calculate stats of un-normalized data / normalized data (mean and standard deviation)

Script: 1_dataConversion.R
Source files: 1_1_shRNAmatched
Output files:
1_2_normtoall_ZP
1_2_normtocontrol_ZP/Amt
1_2_normtocontrol_ZP/Amt

2. Per gene shRNA effect for each marker (by t-test)

  • 2.1 Test consistency of effect of shRNAs targeting the same gene for each marker
    • 2.1.1 Calculate p-value of shRNAs targeting the same gene v.s. all samples with two sample t-test (<- from percentile calculated in last step)
    • 2.1.2 Calculate average Z score and percentile of shRNA targeting the same gene
  • 2.2 Compile data into spreadsheets to summarize average Z score, percentile and p-value of each gene for different markers
  • 2.3 Plot gene average Z-score heatmaps for each marker

Script: 2_convertedDataAnalysis.R
Source files: 1_2_normtocontrol_ZP/Sep
Output files:
2_0_t-test_by_gene/0_Amt_sep
2_0_t-test_by_gene/0_AmtGFP_sep
2_0_t-test_by_gene/1_Amt_compile
2_0_t-test_by_gene/1_AmtGFP_compile

3. Per gene shRNA effect for all markers (by t-sne)

  • 3.1 Cluster shRNAs by Z-score for each marker with t-sne clustering
  • 3.2 Examine clustering of shRNAs targeting the same gene with the t-sne clustering result

Script: 3_Scikit_learn_byshRNA.py; 3_tsne_cbs-plot_byshRNA.R
Source files: 1_2_normtocontrol_ZP/Compiled_*
Output files:
1_2_normtocontrol_ZP/Compiled_* /per*_ clusterbyshRNA (per*: perplexity)

4. Gene knockdown effect clustering (by t-sne)

  • 4.1 Cluster genes by average Z-score for each marker with t-sne clustering (per*: perplexity)
  • 4.2 Plot genes which significantly up/down regulate markers when knocked down (by marker)
  • 4.3 Cluster extraction by louvain method; pathway analysis of extracted clusters

Script: 3_Scikit_learn_byGene.py; 3_tsne_cbs-plot_byGene.R; 4_Tsne_cluster_extraction.R
Source files: 2_0_t-test_by_gene/1_Amt_compile
Output files:
t-SNE coordinates
0. Bubble plot source
1. Annotated bubble plots
1. Non-annotated bubble plots
2. Cluster extraction and plots
3. Extracted cluster pathway analysis