InfernoRDN

InfernoRDN can perform various downstream analyses on large scale datasets from proteomics and microarrays.

Some of the features included with InfernoRDN:

• A set of diagnostic plots (Histograms, boxplots, correlation plots, qq-plots, peptide-protein rollup plots, MA plots, PCA plots, etc).
• Log transforming.
• Rolling up to proteins (3 methods are available).
• LOESS normalization
• Linear Regression Normalization
• Mean Centering
• Median Absolute Deviation (MAD) Adjustment across datasets
• Quantile Normalization
• Principal Component Analysis
• Partial Least Squares Analysis
• ANOVA (multi-way, unbalanced, random effects)
• Heatmaps with Hierarchical and K-means cluster options

InfernoRDN is an updated version of Inferno (which used StatconnDCOM). It supersedes all previous DAnTE (Data Analysis Tool Extension), DanteR, and Inferno versions.

InfernoRDN uses R.NET (https://github.com/jmp75/rdotnet) to communicate with R.

Installation

1. Download and install the latest version of R 4.x

   • https://cran.r-project.org/bin/windows/base/
   • If you prefer R 3.x, you must have R 3.6 or newer

2. Download the InfernoRDN installer from:

   • https://github.com/PNNL-Comp-Mass-Spec/InfernoRDN/releases

3. Run the installer, InfernoRDNSetup.exe

4. Start InfernoRDN using the Start Menu or desktop shortcut

   • You may need to run InfernoRDN as an administrator; see step 6 below

5. The InfernoRDN splash screen will appear and status messages will be shown

   • If a Dialog box appears asking "Would you like to use a personal library?" you should answer Yes to that question, and Yes to the next question regarding the folder to use for the personal library
   • Following this, several Bioconductor packages will be downloaded

6. Diagnosing Startup or Plotting Errors

   • If you see the error message "R failed to install required packages", try manually installing the packages
     • See section "Manual Package Installation" below
     • After installing the packages, use the commands in the "Package Verification" section
     • Provided you can manually load the packages with the library() command, you can likely ignore the "R failed" message shown when InfernoRDN starts
   • Alternatively, you could try running InfernoRDN as an administrator, but this shouldn't be required
     • Right click the shortcut to InfernoRDN and choose "Run as Administrator"
     • Alternatively, navigate to C:\Program Files\InfernoRDN, right click Inferno.exe and choose "Run as administrator"
   • When diagnosing errors, examine the newest rcmd log file at %AppData%\Inferno
     • For example, C:\Users\d3l243\AppData\Roaming\Inferno\rcmdlog.txt
     • or C:\Users\d3l243\AppData\Roaming\Inferno\rcmdlog5.txt

7. Test loading a data file

   • Choose File, Open, Expression File
   • Navigate to C:\Program Files\InfernoRDN\Sample_Data_Files
   • Select SampleInput4DAnTE.csv and click Open
   • Choose column Mass_Tag_ID then click the ">>" button to the left (and just below) "Unique Row ID"
   • Enable checkbox "Protein ID"
   • Select column MinOfOrf then click the ">>" button to the left (and just below) "Protein ID"
   • Select data columns P10A through P19B then click the ">>" button to the left (and below) "Data Columns"
   • Click OK

8. Test the plotting

   • Choose Plot, Correlation
   • Enable checkbox Toggle All, then click OK

Dependencies

InfernoRDN depends on the following:

1. Windows 7 (or newer) with the .NET 4.6 framework or newer

   • https://www.microsoft.com/en-us/download/details.aspx?id=53344

2. R Statistical Environment, version 3.6 or newer

   • https://cran.r-project.org/bin/windows/base/

3. Bioconductor

   • This should get installed automatically by InfernoRDN
   • You may need to manually upgrade Bioconductor, using the steps outlined at https://bioconductor.org/install/

R Packages

InfernoRDN uses the following R packages (from https://cran.r-project.org/):

• amap: Another Multidimensional Analysis Package
• car: Companion to Applied Regression
• lattice: Linear and Nonlinear Mixed Effects Models
• nlme: Linear and Nonlinear Mixed Effects Models
• outliers: Tests for outliers
• fpc: Fixed point clusters, clusterwise regression and discriminant plots
• pls: Partial Least Squares Regression (PLSR) and Principal Component Regression (PCR)
• MASS: Main Package of Venables and Ripley's MASS
• e1071: Misc Functions of the Department of Statistics (e1071), TU Wien
• ggplot2: Various R programming tools for plotting data
• ellipse: Functions for drawing ellipses and ellipse-like confidence regions
• plotrix: Various plotting functions
• scatterplot3d: 3D Scatter Plot
• colorspace: Colorspace Manipulation
• Hmisc: Harrell Miscellaneous
• Cairo: R graphics device using cairo graphics library

Legacy packages (not available for R 3.x or 4.x)

• impute: Imputation for microarray data
• qvalue: Q-value estimation for false discovery rate control

The packages will be installed to either the library folder in C:\Program Files\R\R-3.x.x\library or, more likely (due to permissions) to the R\win-library folder in your "Documents" or "My Documents" folder.

Manual Package Installation

Start R by double clicking R.exe, for example at C:\Program Files\R\R-4.1.1\bin\R.exe

Run this command:

install.packages(c("amap", "car", "lattice", "nlme", "outliers", "fpc", "pls", "MASS", "impute", "qvalue", "e1071", "ggplot2", "ellipse", "plotrix", "scatterplot3d", "colorspace", "jpeg", "Hmisc", "Cairo"), repos='https://cran.revolutionanalytics.com/')

Answer "yes" if prompted with:

• Would you like to use a personal library instead?

Answer "yes" if prompted with:

• Would you like to create a personal library?

Answer "No" when prompted with:

• Do you want to install from sources the packages which need compilation?

Note that package impute is available in e1071

Package Verification

Start R.exe, as described above

Run these commands:

# View installed packages
installed.packages()

# Confirm that packages can be loaded
library(amap)
library(car)
library(lattice)
library(nlme)
library(outliers)
library(fpc)
library(pls)
library(MASS)
library(e1071)
library(ggplot2)
library(ellipse)
library(plotrix)
library(scatterplot3d)
library(colorspace)
library(Hmisc)
library(Cairo)

R Connectivity Issues after Re-install

If InfernoRDN has problems connecting to R after you re-install InfernoRDN (for example, when running ANOVA), try the following

• Exit InfernoRDN
• Navigate to %LocalAppData%\Pacific_Northwest_Nationa
  • For example, C:\Users\d3l243\AppData\Local\Pacific_Northwest_Nationa
  • Yes, the directory name is truncated (it does not end in l)
• Delete any directories that you see there, example names:
  • Inferno.exe_Url_2sg0gwzl52pgvsl5ykzc0musjbmtk3m0
  • Inferno.exe_Url_psrpjex41w0dwbl34gci2qitsop3f50e
• Start InfernoRDN
  • The program will likely re-download all of the required R packages, a process that can take several minutes
• Test the problematic function again

Factor Definitions File

Factors can be defined using the GUI, or by loading a text file listing the factors to associate with each dataset (column name) in the expressions table.

A factor definitions file can be a CSV file (comma-separated) or a .txt file (tab-separated)

The first row of the factor definitions file must have a column named Factor, then column names that match the names in the originally loaded Expressions table. Each subsequent row of the factor file is a new factor name, then the factor value for each dataset.

The following shows example rows of a factor definitions file (tab-separated). There are 6 datasets and two factors (Time and Temperature) defined for each dataset.

| Factor | P10A | P10B | P11A | P11B | P12A | P12B | |-------------|------|------|------|------|------|------| | Time | 0 | 0 | 5 | 5 | 10 | 10 | | Temperature | Hot | Cold | Hot | Cold | Hot | Cold |

Data Files

Inferno saves data as R Session files, with a .dnt extension. These files can be opened with RStudio for custom data analysis.

Importing InfernoRDN Data Into RStudio

• Create a .dnt file inside InfernoRDN using File, Save Session
• Rename the .dnt file to have extension .rdata
• Start R Studio
• Choose File, Open File
• Select the .rdata flie
• Answer "Yes" to the question "Do you want to load the R data file "~/Path/DataFile.rdata" into the global environment?"

The environment tab should now show one or more data matrices

| Variable Name | Description | |---------------|-------------| | Eset | Expression data (primary data loaded into InfernoRDN) | | logEset | Log transformed data | | ProtInfo | Protein to peptide mapping (provided your input data file had a Protein name column) | | qrollupP | Created by QRollup |

Running InfernoRDN Methods Inside RStudio

• Inside RStudio, choose File, Open File
• Select the desired R script, e.g. <code>Rscripts\Rollup\QRollUp.R</code>
• A new tab should appear with the .R file
  • Click the Source button, which will run a command like this in the Console

source('~/Projects/_CommunityApplications/InfernoRDN/Rscripts/Rollup/QRollUp.R')

• Repeat for any additional required files
  • For example QRollUp.R uses RollupScore.R
• Manually call a method, e.g.

QRollup.proteins(Eset, ProtInfo, 30, 0, 3, FALSE, FALSE)

• View data

View(Eset)
View(oneHitProtNames)

Loading All InfernoRDN Scripts Into RStudio

Use the following steps to load every InfernoRDN script into RStudio

• Start RStudio
• Under