GSE196177 Processing Pipeline

Publication

Cell reports (2022) — PMID 35263585

Dataset

Nanostring based analysis of CBF-AML patients

1. 1

   The output of the assay was analyzed by nSolver 4.0 where the mean of the negative spike-in control was used as the threshold of microRNA detection.

   nSolver v4.0

   $ Bash example

   # nSolver is primarily a GUI-based software for NanoString data analysis.
   # The following command is conceptual and represents the analysis step
   # and the specific parameter setting described.
   # Actual execution would involve using the nSolver 4.0 GUI to load the assay output
   # and configure the detection threshold method.
   
   # Assuming an input file from the NanoString instrument (e.g., RCC file)
   INPUT_ASSAY_OUTPUT="assay_output.RCC"
   OUTPUT_ANALYSIS_DIR="nSolver_analysis_results"
   
   # Create output directory if it doesn't exist
   mkdir -p "${OUTPUT_ANALYSIS_DIR}"
   
   # Conceptual representation of the analysis process within nSolver 4.0
   # The user would typically open nSolver 4.0, import the ${INPUT_ASSAY_OUTPUT},
   # navigate to the analysis settings, and select 'Mean of Negative Spike-in Control'
   # as the threshold for microRNA detection.
   
   echo "--- nSolver 4.0 Analysis Configuration ---"
   echo "Input Assay File: ${INPUT_ASSAY_OUTPUT}"
   echo "Detection Threshold Method: Mean of Negative Spike-in Control"
   echo "Output Directory for Results: ${OUTPUT_ANALYSIS_DIR}"
   echo "
   # Please perform the analysis manually using the nSolver 4.0 GUI with the specified settings."
   

   Copy
2. 2

   With the expression profile table generated, patients were then stratified by miR-130a expression using a median split and a Kaplan-Meier curve was drawn by GraphPad Prism 7.

   GraphPad Prism v7 GitHub

   $ Bash example

   # GraphPad Prism 7 is a GUI-based software for statistical analysis and graphing.
   # The following R script demonstrates the analytical steps (median split and Kaplan-Meier) that would typically be performed in a command-line environment, which GraphPad Prism would then visualize.
   
   # Install necessary R packages if not already installed
   # R -e 'if (!requireNamespace("survival", quietly = TRUE)) install.packages("survival")'
   # R -e 'if (!requireNamespace("survminer", quietly = TRUE)) install.packages("survminer")'
   
   # Assume input file 'expression_data.tsv' contains columns:
   # PatientID, miR-130a_Expression, Survival_Time, Event (0=alive, 1=dead)
   
   Rscript -e '
     library(survival)
     library(survminer)
   
     # Load the expression data
     # Replace "expression_data.tsv" with the actual path to your expression profile table
     data <- read.delim("expression_data.tsv", sep="\t", header=TRUE)
   
     # Stratify patients by miR-130a expression using a median split
     median_expr <- median(data$miR.130a_Expression, na.rm = TRUE)
     data$miR_130a_Group <- ifelse(data$miR.130a_Expression > median_expr, "High", "Low")
   
     # Create a survival object
     surv_object <- Surv(time = data$Survival_Time, event = data$Event)
   
     # Fit Kaplan-Meier survival curve
     fit <- surv_fit(surv_object ~ miR_130a_Group, data = data)
   
     # Draw Kaplan-Meier curve and save to a PDF file (GraphPad Prism would do this interactively)
     pdf("kaplan_meier_miR130a.pdf")
     ggsurvplot(fit,
                data = data,
                pval = TRUE,
                risk.table = TRUE,
                legend.title = "miR-130a Expression",
                legend.labs = c("High", "Low"),
                title = "Kaplan-Meier Curve for miR-130a Expression Stratification")
     dev.off()
   
     # Optionally, print summary statistics
     print(fit)
   '

   Copy View on GitHub
3. 3

   All counted reads are provided in the count matrix.

   RSEM (Inferred with models/gemini-2.5-flash) v1.3.3

   $ Bash example

   # Install RSEM (example using conda)
   # conda install -c bioconda rsem
   
   # --- RSEM quantification command ---
   # This command quantifies gene and isoform expression from aligned reads (BAM format)
   # and generates a count matrix (sample_name.genes.results).
   # Replace '/path/to/aligned_reads.bam' with your input BAM file (e.g., from STAR alignment).
   # Replace '/path/to/RSEM_index' with the path to your RSEM reference index.
   # (The RSEM index is typically built once per reference genome/annotation using 'rsem-prepare-reference').
   # Replace 'sample_name' with a unique identifier for your sample.
   
   rsem-calculate-expression --bam \
                             --paired-end \
                             --num-threads 8 \
                             /path/to/aligned_reads.bam \
                             /path/to/RSEM_index \
                             sample_name
   
   # The gene count matrix will be found in 'sample_name.genes.results'
   # The isoform count matrix will be found in 'sample_name.isoforms.results'

   Copy
4. 4

   Data normalization was conducted within the Ncounter software that used a median cut-off of the negative detection probes

   Ncounter vNot specified (Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Ncounter software is a proprietary platform for NanoString nCounter data analysis, typically operated via a graphical user interface (GUI).
   # Direct command-line execution with bash is not standard for this software.
   #
   # The described normalization process involves:
   # 1. Importing raw nCounter data into the Ncounter Analysis Software.
   # 2. Configuring normalization settings within the GUI.
   # 3. Specifying the use of negative detection probes for background subtraction.
   # 4. Applying a median cut-off method for these negative probes to perform normalization.
   #
   # Example conceptual steps within the Ncounter Analysis Software GUI:
   # - Load data files (e.g., RCC files).
   # - Navigate to 'Normalization' or 'Analysis Settings'.
   # - Select 'Negative Controls' for background correction.
   # - Choose 'Median' as the statistical method for negative control background subtraction.
   # - Apply normalization and generate normalized expression data.

   Copy View on GitHub