tutorial.md

layout	title	zenodo_link	level	questions	objectives	time_estimation	key_points	contributors	follow_up_training
tutorial_hands_on	metaQuantome 1: Data creation	https://doi.org/10.5281/zenodo.4037137	Intermediate	How do I perform functional and taxonomy analysis on metaproteomics data? How can I perform quantitation on metaproteomics data? How do I create inputs that can be used in metaquantome to examine differentially expressed proteins?	A taxonomy, functional and quantitational analysis of metaproteomic mass spectrometry data.	1h	Use dataset collections With SearchGUI and PeptideShaker you can gain access to multiple search engines Learning the basics of SQL queries can pay off	subinamehta timothygriffin pratikdjagtap emmaleith mariecrane	type topic_name tutorials internal proteomics metaproteomics

Introduction

{:.no_toc}

Metaproteomics {% cite Metaproteomics_video %} involves characterization of community level expression of microbial proteins from an environmental or clinical sample. Metaproteomics data {% cite Metaproteomics_community_effort %} {% cite Jagtap2015 %} is primarily used to determine the functional status of the microbiome under study along with its taxonomic composition. The Galaxy-P {% cite Galaxy-P_Metaproteomics %} team published a software suite named metaQuantome { % cite Easterly2019 %} to enable quantitative and statistical analysis and visualization of functional, taxonomic expression as well as functional and taxonomy interaction. metaQuantome leverages peptide level quantitative information to analyze the taxonomic, functional expression within the microbial community in different conditions.

{: width="75%"}

metaQuantome offers differential abundance analysis, principal components analysis, and clustered heat map visualizations, across multiple experimental conditions. metaQuantome, an open source tool, is available via command line and also accessible via Galaxy platform for reproducible analysis. As a first step for metaQuantome analysis, metaproteomics data needs to be made compatible for subsequent analysis. With this in mind, we have developed a metaQuantome data generation workflow tutorial that will help users generate inputs for metaQuantome analysis.

{: width="75%"}

To demonstrate the use of the data creation workflow, we have used a thermophilic biogas reactor dataset wherein municipal food waste and manure is digested to generate methane gas. After one round in the reactor, the microbial community was simplified and enriched via serial dilution. This inoculum was then transferred to a solution of cellulose from Norwegian Spruce and incubated at 65°C. Triplicate samples were taken in a time series from 0 to 43 hours after inoculation and mass spectrometry data was acquired on a Q-Exactive (Thermo) mass spectrometer. For this training, we have chosen two time points-8 hour and 33 hour.

{: width="75%"}

Agenda

In this tutorial, we will cover:

1. TOC {:toc}

{: .agenda}

Pretreatments

The first step in a tutorial is to get the data from the zenodo link provided and making sure that it is in the correct format.

Get data

{% icon hands_on %} Hands-on: Data upload

1. Create a new history for this tutorial and give it a meaningful name

   {% snippet faqs/galaxy/histories_create_new.md %} {% snippet faqs/galaxy/histories_rename.md %}

2. Import the files: 6 MZML files, a Protein FASTA file, and an Experimental Design file from [Zenodo]({{ page.zenodo_link }}) or from the shared data library (GTN - Material -> {{ page.topic_name }} -> {{ page.title }})

   https://zenodo.org/record/4037137/files/ExperimentalDesign.tsv
   https://zenodo.org/record/4037137/files/ProteinDB_cRAP.fasta
   https://zenodo.org/record/4037137/files/T2_A1.mzml
   https://zenodo.org/record/4037137/files/T2_B1.mzml
   https://zenodo.org/record/4037137/files/T7A_1.mzml
   https://zenodo.org/record/4037137/files/T7B_1.mzml
   

   {% snippet faqs/galaxy/datasets_import_via_link.md %} {% snippet faqs/galaxy/datasets_import_from_data_library.md %}

3. Rename the datasets (If needed)

4. Check that the datatype ( Make sure they are in the correct formats). 6 MZML files (format=mzml, a Protein FASTA file (format=fasta), and an Experimental Design file (format=tabular)

   {% snippet faqs/galaxy/datasets_change_datatype.md datatype="datatypes" %}

5. Add to each database a tag corresponding to the name of the input data (optional).

6. Build a Dataset list for the four mzml files.

   • Click the Operations on multiple datasets check box at the top of the history panel

   {% snippet faqs/galaxy/collections_build_list.md %}

7. Rename the dataset collection as MZML dataset collection.

{: .hands_on}

Match peptide sequences

For this, the sequence database-searching program called SearchGUI will be used. The created dataset collection of the four MZML files in the history has to be first converted to MGF to be used as the MS/MS input.

Convert mzml to MGF with msconvert

msconvert is used in order to convert the input file type, a mzml data collection, to a mgf file type. The mgf file type can then be used as the Input Peak Lists when running SearchGUI.

{% icon hands_on %} Hands-on: mzml to MGF

1. {% tool msconvert %} with the following parameters:

   • {% icon param-collection %} "Input unrefined MS data": MZML dataset collection
   • "Do you agree to the vendor licenses?": Yes
   • "Output Type": mgf
   • In "Data Processing Filters":
     • "Apply peak picking?": Yes
     • "Apply m/z refinement with identification data?": No
     • "(Re-)calculate charge states?": No
     • "Filter m/z Window": No
     • "Filter out ETD precursor peaks?": No
     • "De-noise MS2 with moving window filter": No
   • In "Scan Inclusion/Exclusion Filters":
     • "Filter MS Levels": No
   • In "General Options":
     • "Sum adjacent scans": No
     • "Output multiple runs per file": No

   {% icon comment %} Comment

   This is a critical step for running this workflow. {: .comment}

{: .hands_on}

{% icon question %} Questions

1. Why do we need to convert the files to MGF?
2. Can we use any other input format?

{% icon solution %} Solution

1. The files have to be converted to MGF for this workflow because we use SearchGUI as the searching tool and it can only read MGF files.
2. Yes, we can also use RAW files as input and just convert RAW files to MGF.

{: .solution}

{: .question}

Search GUI

SearchGUI is a tool that searches sequence databases on any number of MGF files. In this case, the previously made collection of three MGF files (entitles MGF files) will be used as the MS/MS input. This tool will produce an output file, called a SearchGUI archive file. This file will serve as in input for the next tool used, PeptideShaker.

{% icon hands_on %} Hands-on: Search sequence databases

1. {% tool Search GUI %} with the following parameters:

   • {% icon param-file %} "Protein Database": ProteinDB_cRAP.fasta
   • {% icon param-file %} "Input Peak Lists (mgf)": output (output of msconvert {% icon tool %})
   • In "Search Engine Options":
     • "DB-Search Engines": X!tandem
   • In "Protein Digestion Options":
     • "Digestion": Trypsin
     • "Maximum Missed Cleavages": 2
   • In "Precursor Options":
     • "Fragment Tolerance": 0.2
     • "Maximum Charge": 6
   • In "Protein Modification Options":
     • "Fixed Modifications": Carbamidomethylation of C
     • "Variable Modifications": Oxidation of M, Acetylation of Protein N-term
   • In "Andvanced Options":
     • "SearchGUI Options": Default
     • "X!Tandem Options": Advanced
       • "X!Tandem: Quick Acetyl": No
       • "X!Tandem: Quick Pyrolidone": No
       • "X!Tandem: Maximum Valid Expectation Value": 100.0
       • "X!Tandem peptide model refinement": Don't refine
     • "OMSSA Options": Default
     • "MSGF Options": Default
     • "MS Amanda Options": Default
     • "TIDE Options": Default
     • "MyriMatch Options": Default
     • "Comet Options": Default
     • "DirectTag Options": Default
     • "Novor Options": Default

   {% icon comment %} Comment

   Note that sequence databases used for metaproteomics are usually much larger than the excerpt used in this tutorial. When using large databases, the peptide identification step can take much more time for computation. In metaproteomics, choosing the optimal database is a crucial step of your workflow, for further reading see Timmins-Schiffman et al (2017). To learn more about database construction in general, like integrating contaminant databases or using a decoy strategy for FDR searching, please consult our tutorial on Database Handling. {: .comment}

{: .hands_on}

{% icon question %} Questions

1. How many Search Engines can be used?
2. Can the parameters be manipulated?

{% icon solution %} Solution

1. There are 8 database search algorithms, you can use as many as you want. Ideally, 4 database algorithms gives the best results.
2. Yes, The parameters can be manipulated according to the experimental design of the datasets.

{: .solution}

{: .question}

Peptide Shaker

PeptideShaker is a post-processing software tool that processes data from the SearchGUI software tool. PeptideShaker is a search engine for interpretation of proteomics identification results from multiple search engines, currently supporting X!Tandem, MS-GF+, MS Amanda, OMSSA, MyriMatch, Comet, Tide, Mascot, Andromeda and mzIdentML. More specifically, PeptideShaker processes data from the SearchGUI tool through the organization of Peptide-Spectral Matches (PSMs) generated. In addition to organization, it provides an assessment of confidence of the data and generates outputs that can be visualized by users to interpret the results.

{% icon hands_on %} Hands-on: Interpretation of SearchGUI

1. {% tool Peptide Shaker %} with the following parameters:

   • {% icon param-file %} "Compressed SearchGUI results": searchgui_results (output of Search GUI {% icon tool %})
   • "Specify Advanced PeptideShaker Processing Options": Advanced Processing Options
     • "The PTM probabilistic score to use for PTM localization": A-score
   • "Specify Advanced Filtering Options": Advanced Filtering Options
     • "Maximum Peptide Length": 60
   • "Specify Contact Information for mzIdendML": GalaxyP Project contact (Not suitable for PRIDE submission)
   • In "Exporting options":
     • "Creates a mzIdentML file": Yes
     • "Compress results into a single zip file": No
     • "Reports to be generated": PSM Report, Peptide Report, Protein Report, Certificate of Analysis,

   {% icon comment %} Comment

   There are a number of choices for different data files that can be generated using PeptideShaker. A compressed file can be made containing all information needed to view the results in the standalone PeptideShaker viewer. A mzidentML file can be created that contains all peptide sequence matching information and can be utilized by compatible downstream software. Other outputs are focused on the inferred proteins identified from the PSMs, as well as phosphorylation reports, relevant if a phosphoproteomics experiment has been undertaken. More detailed information on peptide inference using SearchGUI and PeptideShaker can be found in our tutorial on Peptide and Protein ID. {: .comment}

{: .hands_on}

Removing Contaminants

{% icon hands_on %} Hands-on: Remove contaminants from PSM report

This Select tool is used to remove all the contaminants from the Peptide Spectral Match (PSM) search results.

1. {% tool Select %} with the following parameters:

   • {% icon param-file %} "Select lines from": output_psm (PSM report output of Peptide Shaker {% icon tool %})
   • "that": NOT Matching
   • "the pattern": con_

2. Rename {% icon galaxy-pencil %} to output file to PSM_Report_no_contaminants

   {% snippet faqs/galaxy/datasets_rename.md %} {: .hands_on}

{% icon hands_on %} Hands-on: Removing contaminants from Peptide report

1. {% tool Select %} with the following parameters:

   • {% icon param-file %} "Select lines from": output_peptides (Peptide Report output of Peptide Shaker {% icon tool %})
   • "that": NOT Matching
   • "the pattern": con_

2. Rename {% icon galaxy-pencil%} the output file to Peptide_Report_no_contaminants

   {% icon comment %} Comment

   In Proteomics, contamination is generally detected as peaks in spectra that did not originate from the samples and can be introduced in the sample from a variety of environmental sources or human error. Identification of these contaminants is critical to enable their removal before data analysis, mainly, to maintain the validity of conclusions drawn from statistical analyses. Thus, this selection tool helps us remove the contaminants that were identified in the spectral data. {: .comment}

{: .hands_on}

{% icon question %} Questions

1. Why is removing contaminants important?

{% icon solution %} Solution

1. Ideally, we would like to remove known contaminants from our samples just to maintain discovering novel proteoforms in our sample.

{: .solution}

{: .question}

Removing file extensions for Quantification

This is a data manipulation step to make the data compatible with other downstream processing tools. The Replace text tool replaces the .mgf extension from the PSM report so that it can be used as an input for FlashLFQ.

{% icon hands_on %} Hands-on: Removing file extensions

1. {% tool Replace Text in a specific column %} with the following parameters:

   • {% icon param-file %} "File to process": PSM_Report_no_contaminants (output of Select {% icon tool %})
   • In "Replacement":
     • {% icon param-repeat %} "Insert Replacement"
       • "in column": Column: 10
       • "Find pattern": .mzml.mgf

2. Rename Input_for_FlashLFQ.

   {% icon comment %} Comment

   Replace Text searches given columns and finds and replaces patterns provided by the user. This tool is removing the extensions (.raw,.mzml,.mgf) in the spectral file column provided by the PeptideShaker tool. This step is critical for FlashLFQ to work. {: .comment}

{: .hands_on}

Extracting Peptide list

{% icon hands_on %} Hands-on: Selecting peptide list

This step selects the peptide column from the Select output ( where we have removed the contaminants)

1. {% tool Cut %} with the following parameters:

   • "Cut columns": c6
   • {% icon param-file %} "From": Peptide_Report_no_contaminants (output of Select {% icon tool %})

2. Rename file as peptide_list. {: .hands_on}

Peptide Quantification

In this tutorial, we are using FlashLFQ as the quantitation tool. The user can choose to work with other quantitation tools, e.g. moFF and MaxQuant are available in Galaxy.

FlashLFQ

FlashLFQ can quantify MS peaks in order to find the abundances of peptides. FlashLFQ is a fast label-free quantification algorithm. Additionally, the abundances of peptides within the sample can be compared between samples as further analysis beyond this workflow.

{% icon hands_on %} Hands-on: Quantification

1. {% tool FlashLFQ %} with the following parameters:

   • {% icon param-file %} "identification file": Input_for_FlashLFQ (output of Replace Text {% icon tool %})
   • {% icon param-collection %} "spectrum files": MZML dataset collection
   • "match between runs": Yes
   • "Use experimental design for normalization or protein fold-change analysis": Yes
     • {% icon param-file %} "ExperimentalDesign.tsv": ExperimentalDesign.tsv
     • "Perform Bayesian protein fold-change analysis": Yes
       • "control condition for Bayesian protein fold-change analysis": S1

   {% icon comment %} Comment

   FlashLFQ is a label-free quantification tool for mass-spectrometry proteomics. It supports both .mzML and Thermo .raw file formats. To run FlashLFQ on Galaxy, there are three main input files:

   • PSM report from Peptide Shaker (Input_for_FlashLFQ)
   • MZML/ RAW spectrum files (MZML dataset collection)
   • Experimental Design File ( ExperimentalDesign.tsv) The Experimental Design file should be a tabular file with a "File", "Condition", "Sample", "Fraction", and "Replicate" column. The "File" column should match your mzml spectrum file name. {: .comment}

{: .hands_on}

{% icon question %} Questions

1. Can FlashLFQ be used with fractionated data?

2. Does FlashLFQ perform peptide and protein level quantification?

{% icon solution %} Solution

1. Yes, FlashLFQ can be used with fractionated datasets and multiple conditions

2. FlashLFQ performed both peptide level and protein level quantification. For protein level, FlashLFQ used Bayesian Fold change analysis.

{: .solution}

{: .question}

Filtering peptides that are less than 50 amino acids

{% icon hands_on %} Hands-on: Extracting peptides<50 amino acids

This is a data manipulation tool. Here, we select those peptides with less than 50 amino acids in length.

1. {% tool Filter data on any column using simple expressions %} with the following parameters:

   • {% icon param-file %} "Filter": peptide_list (output of Cut {% icon tool %})
   • "With following condition": len(c1)<=50
   • "Number of header lines to skip": 1

2. Rename as Unipept_peptide_list.

   {% icon comment %} Comment

   Unipept fails with peptides more than 50 amino acids in length, thus we decided to work with peptides that are less than 50 amino acids.

   {: .comment}

{: .hands_on}

Manipulating text for metaQuantome

{% icon hands_on %} Hands-on: Text manipulation for metaQuantome intensity file

Regex Find And Replace goes line by line through the input file and will remove any patterns specified by the user and replace them with expressions also specified by the user. In this case, Regex Find And Replace is being used on a FlashLFQ output file and manipulating the header to make it compatible with metaQuantome along with completely removing the N-terminus and C-terminus tag in the peptide sequences.

1. {% tool Regex Find And Replace %} with the following parameters:

   • {% icon param-file %} "Select lines from": QuantifiedPeptides (output of FlashLFQ {% icon tool %})
   • In "Check":
     • {% icon param-repeat %} "Insert Check"
       • "Find Regex": Base Sequence
       • "Replacement": peptide
     • {% icon param-repeat %} "Insert Check"
       • "Find Regex": Intensity_
     • {% icon param-repeat %} "Insert Check"
       • ”Find Regex”: NH2-
       • ”Replacement”: leave it blank
     • {% icon param-repeat %} "Insert Check"
       • ”Find Regex”: -COOH
       • ”Replacement”: leave it blank

2. Rename the file as Intensity

{: .hands_on}

Functional and Taxonomy annotation

Unipept for taxonomy annotation

Unipept {% cite Mesuere2018 %} is used again to match tryptic peptides and find the taxonomy and lowest common ancestor of each peptide.

{% icon hands_on %} Hands-on: Taxonomy annotation

1. {% tool Unipept %} with the following parameters:

   • "Unipept application": pept2lca: lowest common ancestor
     • "Equate isoleucine and leucine": Yes
     • "allfields": Yes
   • "Peptides input format": tabular
     • {% icon param-file %} "Tabular Input Containing Peptide column": Unipept_peptide_list (output of Filter {% icon tool %})
     • "Select column with peptides": c1
   • "Choose outputs": Select all

   {% icon comment %} Comment

   There are two Unipept in this workflow, One for taxonomy and other for function. {: .comment}

{: .hands_on}

{: width="75%"}

{% icon question %} Questions

1. Can any other taxonomy and functional tool be used apart from Unipept?

{% icon solution %} Solution

1. Yes, any tool can be used for taxonomy and functional output. Please make sure the output has the information that includes peptide,taxon_name, taxon_id, genus, species etc.

{: .solution}

{: .question}

The JSON output from the Taxonomy can be visualized using the visualize option and Select the Unipept Taxonomyviewer.

{: width="30%"}

{: width="30%"}

{: width="75%"}

Unipept for Functional annotation

Unipept is used to match tryptic peptides and find the taxonomy and Functional annotation of the peptides. Unipept is used to match sample tryptic peptides to proteins using a fast-matching algorithm. Although Unipept can be accessed and used through the web page, the use of Unipept on Galaxy allows the production of output datasets including the peptide information to be used in sequential steps. Unipept requires a list containing the peptide sequences which was generated by Query Tabular.

{% icon hands_on %} Hands-on: Functional annotation

1. {% tool Unipept %} with the following parameters:

   • "Unipept application": peptinfo: Tryptic peptides and associated EC and GO terms and lowest common ancestor taxonomy
     • "Equate isoleucine and leucine": Yes
     • "retrieve extra information": Yes
     • "group responses by GO namespace (biological process, molecular function, cellular component)": Yes
     • "allfields": Yes
   • "Peptides input format": tabular
     • {% icon param-file %} "Tabular Input Containing Peptide column": Unipept_peptide_list (output of Filter {% icon tool %})
     • "Select column with peptides": c1
   • "Choose outputs": Select all

   {% icon comment %} Comment

   There are two Unipept in this workflow, One for taxonomy and other for function. Please select all the output options from Unipept. {: .comment}

The JSON output from the Taxonomy can be visualized using the visualize option and Select the Unipept Taxonomyviewer.

{: .hands_on}

Extracting EC values

{% icon hands_on %} Hands-on: Extract EC numbers

The cut tool cuts out specific columns from the dataset. In this case, the cut tool is being used to extract columns 1 (peptide) and 3 (EC number) from the dataset peptinfo EC.tsv output. This is a manipulation tool for metaQuantome's convenience.

1. {% tool Cut %} with the following parameters:
   • "Cut columns": c1,c3
   • {% icon param-file %} "From": output_ec_tsv (output of Unipept {% icon tool %})
2. Rename EC_table. {: .hands_on}

Filtering confident peptides

Query Tabular is a tool that can load tabular data into a SQLite database. This step precedes UniPept, as a list containing the peptide sequences must be generated. In this step a list of gene ontology (GO) terms is being generated.

{% icon hands_on %} Hands-on: Filtering confident peptides

1. {% tool Query Tabular %} with the following parameters:
   • In "Database Table":
     • {% icon param-repeat %} "Insert Database Table"
       • {% icon param-file %} "Tabular Dataset for Table": output_go_tsv (output of Unipept {% icon tool %})
       • In "Table Options":
         • "Specify Name for Table": Goterm
         • "Specify Column Names (comma-separated list)": peptide,total_protein_count,go_term,protein_count,go_name,go_funct
   • "SQL Query to generate tabular output":

SELECT Goterm.*
FROM Goterm
WHERE ((1.0*Goterm.protein_count)/(1.0*Goterm.total_protein_count)) >= 0.05

• "include query result column headers": Yes

2. Rename as Unipept_Function.

   {% icon comment %} Comment

   In the Unipept API output, the threshold is set to 0.5% of the overall number of peptides unambiguously assigned to a taxon at a particular taxonomic rank level. Here in the Galaxy platform, we are using Query tabular to perform this filtering. {: .comment}

{: .hands_on}

Removing Hashtag from output

This step is to remove the hashtag from the Peptide header in the Unipept output.

{% icon hands_on %} Hands-on: Remove # from peptide header

1. {% tool Replace Text %} with the following parameters:

   • {% icon param-file %} "File to process": Unipept_Function (output of Unipept {% icon tool %})
   • In "Replacement":
     • {% icon param-repeat %} "Insert Replacement"
       • "Find pattern": #peptide
       • "Replace with:": peptide

2. Rename file as All_functions {: .hands_on}

Filter - EC values

We are using this Query tabular to rename the output that we obtained from the Cut column tool.

{% icon hands_on %} Hands-on: Extracting EC for metaQuantome

1. {% tool Query Tabular %} with the following parameters:

   • In "Database Table":
     • {% icon param-repeat %} "Insert Database Table"
       • {% icon param-file %} "Tabular Dataset for Table": EC_table (output of Cut {% icon tool %})
       • In "Table Options":
         • "Specify Name for Table": ec
         • "Specify Column Names (comma-separated list)": peptide,go_ec
   • "SQL Query to generate tabular output": SELECT * FROM ec
   • "include query result column headers": Yes

2. Rename file as go_ec {: .hands_on}

Filter - Biological Functions

The filter tool allows restriction of the dataset using simple conditional statements. This step is used to filter out the GO terms with biological processes and the corresponding number of peptides associated with these terms.

{% icon hands_on %} Hands-on: Extracting biological processes for metaQuantome

1. {% tool Filter %} with the following parameters:

   • {% icon param-file %} "Filter": Unipept_Function (output of Query Tabular {% icon tool %})
   • "With following condition": c5=='biological process'
   • "Number of header lines to skip": 1

2. Rename file as go_bp {: .hands_on}

{: width="60%"}

Filter - Cellular components

This step is used to filter out the GO terms with cellular components and the corresponding number of peptides associated with these terms.

{% icon hands_on %} Hands-on: Extracting cellular component for metaQuantome

1. {% tool Filter %} with the following parameters:

   • {% icon param-file %} "Filter": Unipept_Function (output of Query Tabular {% icon tool %})
   • "With following condition": c5=='cellular component'
   • "Number of header lines to skip": 1

2. Rename file as go_cc {: .hands_on}

{: width="60%"}

Filter - Molecular Function

This step is used to filter out the GO terms with molecular function and the corresponding number of peptides associated with these terms.

{% icon hands_on %} Hands-on: Extracting molecular function for metaQuantome

1. {% tool Filter %} with the following parameters:
   • {% icon param-file %} "Filter": Unipept_Function (output of Query Tabular {% icon tool %})
   • "With following condition": c5=='molecular function'
   • "Number of header lines to skip": 1
2. Rename file as go_mf {: .hands_on}

{: width="60%"}

Conclusion

{:.no_toc}

This completes the walkthrough of the metaQuantome data creation workflow. This tutorial is a guide to have datasets that are metaQuantome ready/compatible and can be used for metaproteomics research. We have incorporated only two conditions in this workflow but users can use as many as they want. Researchers can use this workflow with their data also, please note that the tool parameters and the workflow will be needed to be modified accordingly.

This workflow was developed by the Galaxy-P team at the University of Minnesota. For more information about Galaxy-P or our ongoing work, please visit us at galaxyp.org

{: .comment}