Proturn

A proteomics application for analyzing 2H2O labeling data

Overview

ProTurn v.2.1.01 is a custom proteomics application written to facilitate the analysis of deuterium oxide (heavy water) labeling mass spectrometry data.

ProTurn currently supports Thermo LTQ-Orbitrap/ XL/ Velos/ Elite instruments and protein identification workflows using SEQUEST, ProLuCID, or MaxQuant.

ProTurn requires Java 8 or above, on Microsoft Windows and Macintosh OS X. The Java SE Runtime Environment 8 package can be downloaded from Oracle. Note that for large data sets it may be necessary to increase the memory heap size that is allocated by the operation system to the application. To increase the heap size, launch ProTurn in the terminal (OS X) or command prompt (Windows) with the command

  [java] -Xmx6144m -jar ProTurn.jar

to specify 6144 MB of memory, where [java] denotes the path of the Java installation on the system.


File Input

For mass spectral data, ProTurn currently accepts [.mzML] raw spectrum files generated by the default command line settings of MSConvert (no compression, [.mzML] format, 64-bit m/z precision, 32-bit intensity precision). For database search results, ProTurn currently accepts four common protein identification workflows:

  1. ProLuCID and DTASelect as implemented in the Integrated Proteomics Pipeline,
  2. SEQUEST as implemented in Thermo BioWorks v.3.3 (or comparible) combined with optional protein filtering by Proteome Software Scaffold v.3 (or compatible),
  3. SEQUEST HT as implemented in Thermo Proteome Discover v.1.4 (or compatible).
  4. MaxQuant/Andromeda v.1.3.05 (or compatible).

To start analysis, create an analysis folder in the desired location on the hard drive. Create a subfolder inside the main analysis folder called [/mzml] to host the raw files. Inside the [/mzml] folder, create further subfolders named after the sample labeling time points in the format of [h]/[d][##], e.g.,[h12], [d00], [d02], to denote 12 hours, 0 days, 2 days, etc. Put all the mass spectrum files belonging to a particular time point into the folder. Note that the file names must be the same as referred to in the search result files for them to be correctly recognized.

To run ProTurn from the ProLuCID/ DTASelect workflow, export the DTASelect search result files in text format. Rename the search results in the format of [h/d##.txt], e.g., [h12.txt], [d00.txt], [d07.txt], corresponding to the experimental time points and the subfolders inside the [/mzml] folder where the spectra are stored.

To run ProTurn from the BioWorks workflow, open the SEQUEST search result [.srf] files in Thermo BioWorks v. 3.3 or another compatible version. The peptide search results must be exported from BioWorks as [.xml] files and placed into respective subfolders named after the time points ([/d00], [/d00], etc.) inside a manually created [/srf-xml] subfolder.

To create the [.xml] files from the individual SEQUEST search result [.srf] files, open the [.srf] file in BioWorksBrower. Right click anywhere on the result view, then select Export > XML with the Include All Proteins option. Doing so will export every peptide in the current view. To export only the confidently identified peptides, it is recommended that the view is first filtered with appropriate parameters. To do so, go to Display > Options > Filter & Sort. Recommended filter types include Peptide Delta CN: 0.100, Peptide Xcorr: 1.50, 2.00, 2.50, 0.00, and Number of Distinct Peptides: 2.

When reading from [.srf-xml] files, ProTurn can incorporate optional protein redundancy filter results from Proteome Software Scaffold v.3 or other compatible versions. To do so, create a [/scaffold] subfolder in the analysis folder prior to running ProTurn. Open the Scaffold result [.sfd] files for the project in Scaffold v.3 and then export them by selecting Export > To Excel > Peptide reports. Rename the resulting [.xls] files of different labeling time points accordingly, e.g., [d00.xls], and place them inside the [/scaffold] subfolder.

To run ProTurn using the Thermo Proteome Discoverer SEQUEST HT workflow, create a [/msf] subfolder inside the main analysis folder. Open the Proteome Discoverer SEQUEST search result [.msf] file with Proteome Discoverer 1.4 or other compatible versions. Apply appropriate filters in the Result Filters tab. Recommended filters include Peptide Confidence: High and Peptides Per Protein: 2 .

To export the search result files, go to File > Export > To Text (tab delimited). In the pop-up menu, select Criteria > Peptides, and under the Peptides tab, select Best PSMs of all peptide groups , then click Export. Rename the resulting test files according to the labeling time points of the experiment, e.g., [d00.txt], and place them inside the [/msf] subfolder.

To run ProTurn from the MaxQuant/Andromeda workflow, create a main analysis folder as described above. After searching is done, rename each search result folder in the format of [h/d##.txt], e.g., [d00.txt], corresponding to the experimental time points. Each search result folder must contain an [msms.txt] file and a [peptides.txt] file. All other files may remain in the folder, but they are not required. Place the renamed subfolders inside a folder named [/max] inside the main analysis folder.

The MaxQuant/Andromeda directory structure should match the following:

  • /max/d03/peptides.txt
  • /max/d03/msms.txt
  • /max/d06/peptides.txt
  • /max/d06/msms.txt

After the analysis folder and subfolders are populated, go to the ProTurn user interface and specify the location of the analysis folder under the Integrate Peaks tab.

Back to top


Integration

Several options for integration are available in ProTurn to accommodate different instrumentation profiles.

  1. Mass Window: The mass accuracy (in ppm) to integrate in chromatographic space for each mass isotopomer. A permissive 200-ppm mass window is recommended for Orbitrap instruments running in 7,500 resolution or above, although co-elution peaks may reduce the number of peptides that fit to the model. A mass window of 75 ppm is recommended for 30,000 or 60,000 resolution. Select the appropriate integration peak width based on the resolving power of the experiment.
  2. Peaks: The number of mass isotopomers to integrate for each peptide cluster.
  3. SmoothPoints: How many points before and after each data point to average for boxcar smoothing (0 = no smoothing).
  4. Minimum Area: A minimum peak area filter for excluding low-intensity peaks from integration (default = 0).

Type in the desired parameters and click the Save Parameters button. The program should now have created a [quant.conf] file in the analysis directory. Click the Integrate Peaks button. Go to the Fit Kinetic Curve tab after the console indicates that integration is complete. Integration time varies with the speed of the computer and the size of the dataset. Integration should take several hours for most experiments.

To improve performance on a multi-core machine, change the number of integration threads. The number of threads can be changed only when integration is paused or before starting integration.

Back to top


Curve Fitting

Three fitting options are currently implemented in ProTurn.

  1. NS (Default): The default and recommended nonlinear fitting method is suitable for both steady-state (fast) and non-steady-state (gradual) precursor enrichment. It uses the known values of kp, pss, a, and N to calculate the best-fit value of k to fit the data. It requires the input of heavy water enrichment rate (kp) and level (pss), whereas a and N are calculated automatically from the peptide sequence.
  2. NS (Free): A free-fitting method that performs optimization for all parameters. This is recommended for testing and method development only.
  3. SS: A steady-state enrichment fitting option which uses a simple first-order exponential decay function as previously described (Kim et al., Mol Cell Proteomics 2012). This is suitable for cell culture experiments or small mammal experiments designed to boost precursor enrichment levels to the plateau quickly using heavy water injections, and has the advantage of not requiring kp, pss, a, N, or the plateau A0 value to be known.

Select the desired fitting method, and then input the fitting parameters:

  1. R^2 threshold: This is for the SS curve-fitting method only, and will dictate the R2 value an individual isotopomer fitting must attain before it is considered for protein-level fitting.
  2. No. of min data points: The minimum number of data points in which a peptide and its associated protein must be explicitly identified in the database search result files for it to be considered for fitting. If only one experiment per time point is run, this would be equal to the number of minimum time points option. The minimum should be at least 4.
  3. No. of min time points: The minimum of data points in which a peptide and its associated protein must be explicitly identified in the database search result files for it to be considered for fitting. Curve fitting can be performed with a single time point; in this case a peptide's A0 value will be used as the t0 data point.
  4. % Precursor: This box inputs the pss value of the labeling experiment.
  5. Precursor turnover rate: This box inputs the kp value of the labeling experiment.

Type in the desired parameters and click the Save Parameters button. The program should now have created a [fitter.conf] file in the analysis directory. Click the Fit Kinetic Curve button. Go to the Results tab after the console indicates that fitting is complete. Curve-fitting should take 15-60 minutes for most datasets.

Back to top


Output

The fitting results can be inspected under the Results tab.

Selecting the result directory will allow the result files of any previous analyses to be visualized. The peptides will be displayed in a table that can be sorted and filtered by the number of data points or R2 value. Clicking on a particular peptide and then the Display Graph button will display the individual data points and best-fit curve for the peptide. The raw result is saved in [.out] format in the same directory as the specified analysis folder and can be viewed using a text editor.

Back to top


Contact Us

Visit us at the Ping Lab website. For inquiries, please contact the Ping Lab at proturn@heartproteome.org

Back to top


Download Proturn

To request an evaluation copy of ProTurn for academic use, please email your name, lab, and organization to the Ping Lab at proturn@heartproteome.org

Back to top