Unit tests for SanXoT

We present here three unit tests for windows that can be performed using the following programs in the SanXoT software package:

• Aljamia (data parser)
• Klibrate (weight calibrator)
• SanXoT (integration and variance calculation)
• SanXoTSieve (removal of outliers)
• Cardenio (merger of experiments)
• SanXoTSqueezer (category filter)
• SanXoTGauss (category cumulative gaussian graphs)
• Sanson (category clustering)
• Coordinometer (calculator of the degree of coordination)

For detailed explanations, check: exploring SanXoT features.

Note that links to SanXoT windows standalone executables are provided in this unit test, but download links to the source code and up-to-date versions of SanXoT are available at: SanXoT software package.

Test 1: The fundamental workflow

The fundamental workflow consists of the steps to quantify proteins using PSMs and quantitative information, based on the WSPP (Weighted Spectrum, Peptide and Protein) statistical model, as described, using the following the steps:

• calibration of spectra
• integrating from spectra to peptides
• integrating from peptides to proteins
• quantifying proteins

This test makes use of four programs: Aljamia, Klibrate, SanXoT, and SanXoTSieve.

To run this unit test, follow these steps:

1) download the windows executables, SanXoT.zip. Unzip the whole content in a folder specific for the program.

2) download the files for the unit test, SanXoT_test1.zip. Unzip to a working folder. You should have two folders, one with the input for this test, and another with the expected results. In the former, you will find two text files:

• commands_test1.bat, a windows batch file with command lines to run this sample analysis.
• startingFile.xls, a tab-separated-values text file with identifications and quantitative data from a proteomics experiment.

3) get a) the path of your working folder (where you have unzipped the startingFile.xls), b) the path of the folder where you have unzipped the windows executables, and c) the unit of the latter (C:, D:, etc); modify the following lines in the command_test1.bat file accordingly:

set unit=D:
set programFolder="D:\SSP"
set workingFolder="D:\SanXoT_test1"

4) execute the commands_test1.bat file, copy and paste (or drag and drop) the whole commands_test1.bat text into a command prompt window (you can also just double click the bat file; however, if an error arises, the CMD window created will close immediately, so the text of the error will not be available).

5) wait until it finishes. It should take 30-60 seconds (for a regular PC with 64-bit Windows 10, 3.4 GHz, 32 GB RAM).

6) compare your results with the data in folder SanXoT_test1_results, at SanXoT_test1.zip (for example using a comparison software such as Beyond Compare). Only the files ending in *_log.txt and *_infoFile.txt should have differences (and only due to the different path for files, and the timestamp, all other features being identical, such as variances, Levenberg-Marquardt steps, and options).

Test 2: Experiment merging

Protein quantifications from different technical or biological replicates can be merged into experiment-independent quantifications, giving the biological and technical variance. In this test we

• use the data from Test 1
• perform the fundamental workflow for another parallel experiment
• we merge both experiments

This test makes use of four programs: Klibrate, SanXoT, SanXoTSieve, and Cardenio.

To run this unit test, follow these steps:

1) use the windows executables unzipped in step 1 of Test 1 (SanXoT.zip).

2) download the files for the unit test, SanXoT_test2.zip. This contains:

• all the files generated in Test 1
• the tagFile.txt (for Cardenio)
• the commands_test2.bat.

3) remember to change the following lines in the command_test2.bat file, as in step 3 for Test 1:

set unit=D:
set programFolder="D:\SSP"
set workingFolder="D:\SanXoT_test2"

4) execute the commands_test2.bat file (do the same as in step 4 in Test 1).

5) wait until it finishes. As in Test 1, this test should take 30-60 seconds (for a regular PC with 64-bit Windows 10, 3.4 GHz, 32 GB RAM).

6) as in Test 1, the only files that are not identical should be the merge_logFile.txt from Cardenio, and the *_infoFile.txt from SanXoT, SanXoTSieve and Klibrate.

Test 3: Systems biology

The SanXoT Sowftware Package can be used to perform systems biology analises using the SBT (Systems Biology Triangle) as described. In this test we:

• generate a relations file using data from the DAVID bioinformatics resource (you can download yourself the data from them, but for your convenience we already included it)
• use the data from Test 2
• calculate a category-level fold-change
• make the analysis of the systems biology.

This test makes use of seven programs: Camacho, SanXoT, SanXoTSieve, SanXoTSqueezer, Sanson, SanXoTGauss, and Coordinometer.

To run this test, follow these steps:

1) use the windows executables unzipped in step 1 of Test 1 (SanXoT.zip).

2) download the files for the unit test, SanXoT_test3.zip. This contains:

• all the files generated in Test 1 and Test 2
• the tab-separated text file SB_Homo_19dic-2017.txt (a set of data downloaded from DAVID)
• the text file fastaHeadersNoSequences.fasta (not essential, but important to provide fasta headers as protein descriptors instead of accession numbers, making the results more human readable for the biological interpretation; the whole FASTA file used for the identification can be used, but here we have removed the amino acid sequences for file size reasons)
• the commands_test3.bat.

3) remember to change the following lines in the command_test3.bat file, as in step 3 for Test 1 and Test 2:

set unit=D:
set programFolder="D:\SSP"
set workingFolder="D:\SanXoT_test3"

Additionally, if you want to get the similarity graph the category clustering algorithm (Sanson), you will need:

• to have installed Graphviz (an open source external software, which includes the interpreter for the [1] DOT language, used by Sanson to generate the graphs for the clusters)
• from the program folder, modify the dot.ini file changing the following line to include the path to the bin-folder of Graphviz, which could be different for each user:

dotlocation = C:\Program Files (x86)\Graphviz2.36\bin

4) execute the commands_test3.bat file (do the same as in step 4 in Test 1).

5) wait until it finishes. This test should take 10-20 seconds (for a regular PC with 64-bit Windows 10, 3.4 GHz, 32 GB RAM).

6) as in Test 1 and Test 2, the only files that are not identical should be the *_logFile.txt from Camacho, SanXoTSqueezer, Sanson, SanXoTGauss and the Coordinometer, as well as the *_infoFile.txt from SanXoT and SanXoTSieve. Additionally, if you use different versions of Graphviz (this test used Graphviz 2.36) you might find differences in the file sanson_simGraph.pdf (which is a pdf containing the final graph with the category clusters).