Menu
Selected Models
Search
Quicksearch
Search for Reactions, Metabolites, Genes, and Organisms.
If ticked all models are searched, even not selected ones.
all
       
MEMOSys Help

User Manual

Introduction

This document is a user guide for the web application MEMOSys2. Currently, a public instance is hosted by the ICBI institute and provides free access to several models (Innsbruck Medical University).

Quick start

  • The homepage of the MEMOSys database lists all available publicly available models. You can download the SBML file and (if available) a graphical image of the model directly from this page. Furthermore, you can select which models you want to use in future steps.
  • Use the menu on the left to navigate to Details of Models where you can see the hierarchy of models and where new derived models can be created. The Compare Models site lets you compare two models to get an overview of the differences.
  • Use the Summary menu to get information about reactions, metabolites, and genes.
  • The Search box at the bottom of the menu lets you quickly search for various entities (see section Quicksearch)

MEMOSys

The overall goal of systems biology is the simultaneous study of all processes and dynamic interactions at molecular level in order to draw conclusions which may not be apparent when only individual components are considered. Genome-scale modeling is a promising approach to systematically analyze complex cellular systems. Metabolic models have proven to be valuable for increasing the product yield, predicting the effect of gene deletions, improving gene annotation, and identifying regulatory mechanisms. As models are constructed based on annotated genomes, current advancements in next generation sequencing technologies will foster the development of new models.
Therefore, a bioinformatics platform for the management, storage, and development of metabolic models has been established. The web based Metabolic Model System (MEMOSys) supports the development of new models by providing a built in version control system which offers access to the complete reconstruction history. Moreover, the integrated web board, the fine-grained authorization system, and the definition of user roles allow collaborations across departments and universities. Research on existing models is facilitated by a powerful search system, references to external databases, and a feature-rich comparison mechanism. MEMOSys provides customizable data exchange mechanisms using the SBML format to enable analysis in external tools. The web application is based on the Java EE framework and offers an intuitive user interface. It currently contains several well annotated and public available models.
In summary, the implemented bioinformatics platform provides researchers a novel application facilitating the management and development of metabolic models.

Updates in 2.0

The following updates were introduced in version 2.0:
  • Derived model creation
  • Create a hierarchy of models - see Derived Model
  • New publicly available models
    The public instance of MEMOSys now contains 20 different genome-scale metabolic models
  • Interface improvements
    The design of MEMOSys was changed to remove unnecessary information and provide clear navigation
  • MEMOSys as a virtual machine image
    You can now download MEMOSys as a fully configured virtual machine image from the webpage http://www.icbi.at
  • Annotation improvement feature
    In addition to manually improve the components stored in the MEMOSys database, the system now allows to semi-automatically enhance the annotation of multiple components using SBML files.

Citation

Please cite MEMOSys as follows:
Pabinger S, Rader R, Agren R, Nielsen J, Trajanoski Z. MEMOSys: Bioinformatics platform for genome-scale metabolic models. BMC Syst Biol. 5: 20-20 (2011)

User Guide intention

This document is not meant to be read from the first to the last page. It is more a reference book which can be consulted if you get lost in the system.

Download

If you want to install MEMOSys on your system, consult the information found at:

http://www.icbi.at/MEMOSys

Direct download

The installation description is available below:

MEMOSys Installation Description

The API documentation can be found at:

MEMOSys API documentation

The database schema can be found at:

MEMOSys database schema

General Information

This section describes the general layout and functionality of the web application.

Header Section

The Home link sends you back to the welcome screen (see section Welcome Screen). This is also the screen where you can select models.
The Feedback link provides information where to put your feedback.
The Contact link provides contact information.
The Help link sends you to the help page displaying download and installation information.
The News sends you to the news page.
The About displays information about the system.

At the right side of the application there are three icons that let you change the spatial usage of the browser window:
  • first icon - resizes the window to the default size
  • second icon - stretches the window to the full width of the screen
  • third icon - uses the full window and hides the images in the header section
This setting is stored for users that are logged in into the system and is applied every time when they log.

User Bar

The User Bar is located right beneath the Header Section. It allows registered users to log in into the system. Once they are logged in, the system provides the possibility to change the password and to log out again.

Navigation

Menu
The navigation section is positioned on the left side of the screen. When clicking on a top level menu the submenu is revealed providing links to the respective pages.
Selected Models
Displayed below the Menu section are the currently selected models. By pressing on Selected Models you will be sent back to the welcome screen (see section Welcome Screen).
Quicksearch
Beneath the selected models section is a quick search input field that lets you search for various properties at once (see section Quicksearch).

Selection / deselection of ListBoxes

MEMOSys uses ListBoxes that allow the selection of several items at once throughout the application. To select items click with ctrl hold down on the items that should be selected. To deselect items use the same mechanism as for selecting items. If nothing should be selected simply click on all colored entries with ctrl hold down until nothing is selected anymore.


Fig: Left: click with ctrl hold down on the items that should be selected. Right: to deselect items click with ctrl hold down on the items that should be de-selected. To deselect all items simply click on all colored entries until nothing is selected.


Login

The system allows registered users to login granting them their associated rights.


Fig: Login screen for MEMOSys

Miriam

Links to external resources (databases,...) are stored using the miriam format (www.ebi.ac.uk/miriam). Using this format the webserver can store an arbitrary number of links and can be easily adapted to link to new resources. The miriam identifiers (e.g: urn:miriam:ec-code:4.1.2.27) are transformed into hyperlinks which are then presented to the user:


Fig:Miriam Link

Model

Model list and derived model creation

One way to display currently available models is to use the ModelList. The list gives an overview of all available models that are currently stored in the database. The first column graphically displays the hierarchy of the models. By selecting a model its detailed information is shown. The last two columns show icons for editing the selected model or creating a new derived model. When pressing the latter icon an overflow window is shown where the modelId and name for the derived model need to be specified. After successful creation the new model is shown in the ModelList. Derived models are created in the background and appear in the ModelList once the creation process finished.


Fig: Model list. Mouse points to icon for creating derived models.


Fig: Overflow window for derived model creation.


Fig: New derived model was created (Yeast1).

Model

The model view provides detailed information about a selected model. The lower part shows attached threads and the history of the model (if viewed in a derived model the history of the parent model is displayed as well). In addition, the reaction history (including changes in the parent model) can be displayed. Reactions for this model can be shown under section Reactions, if the model has been selected (see section Welcome Screen).


Fig: History of reactions for selected model.

Model editing

When editing a model with one or several derived models, a checkbox next to each field is displayed. When ticked, the current value is propagated to the derived models overriding the value in each derived model.


Fig: Edit model with derived models.

Derived Model

MEMOSys offers the possibility to create derived models out of existing models. Each derived model is in essence an exact copy of the parent model where the new model contains an association to the parent model and each reaction of the derived model contains a reference to its parent reaction. If a parent reaction is modified, all associated child reactions are updated as well. Furthermore, if a child reaction is modified, it loses its reference to the parent reaction, as it is now considered as an independent reaction.


Part a) displays three GEMs where model iFF708_1 is a derived model of iFF708, and iFF708_2 is a derived model of iFF708_1. Model iFF708_1 and iFF708_2 are copies of model iFF708 and contain exact copies of all reactions present in model iFF708. Part b) shows a use case where reaction 1 was modified in model iFF708_1 (its name was changed to reaction X) resulting in the loss of the reference to reaction 1, and in the update of reaction 1 in model iFF708_2. If reaction 1 in model iFF708 is updated reaction X will not be changed in model iFF708_1 and model iFF708_2.

Access rights

Models in MEMOSys can be marked as either public available or private.
Public available models are visible to all users and display only accepted versions of entities (e.g.: reactions).
Private models are only visible to users that have been granted access to this model. For users of a specific model all versions of reactions, even not accepted ones, are shown.
Each model in MEMOSys is connected to a resource in the usermanagement system. This allows granting fine grained access rights to each model. Resources can be granted to either users or groups. Be careful when editing the model-ID as it changes the identifier of the resource in the usermanagement system which may result in access problems with previous versions of the model. Please contact an administrator when you want to change your model-ID.

Create a new model and assign access rights

When creating a new model in MEMOSys the following steps have to be performed
  1. Create new model in MEMOSys
  2. Log in into the usermanagement system
  3. Create a new resource with exactly the same name as the model-ID (e.g.: nm1234) - grant all rights
  4. Create a new group with the same name as the resource
  5. Create a new ACL (Access control list) for that resource and assign it to the new group or to exisiting users/groups


Fig: Create a new resource in the usermanagement system.


Fig: Assign access control rights to the newly created resource (either group or user).

Welcome Screen

The MEMOSys welcome screen is divided into three sections which can be opened and closed independently:


Fig: Welcome screen of MEMOSys.

  • Select Models - allows users to select models and their version for which reactions should be retrieved. By ticking the box next to the model, the model is automatically selected. The selection is stored in the database for registered users and loaded on their next visit.
  • Latest Changes - displays the latest changes that have been made to reactions and metabolites.
  • Webboard - displays the latest threads of the webboard.

Entities

This chapter describes the entities that are available in MEMOSys. The entries are displayed in lists providing links to pages that display detailed information about the selected entry. Furthermore, given the access rights, users have the possibility to edit and delete items.

Reactions

The reaction list displays all available reactions for the selected models. Reactions can be viewed by pressing on their name/abbreviation. Editors (see section User Classes) are allowed to edit and delete reactions. Only reactions of a model where the latest version is selected are allowed to be edited. In addition, the reaction list directly displays the outcome of the balance check.
Export ReactionList
The result list of the reaction query can be exported either in SBML, PDF, or Excel format. In each case only the result of the query is exported (not all reactions of selected models) and only the selected parameters are included into the export file.


Fig: Export reaction list either into SBML, pdf, or excel format.

Copy Reaction
Reactions can be copied between models (or within the same model) by pressing the Copy Reaction button on the reaction page. After pressing the button an overview window is presented where the user specifies the target model.


Fig: Copy reaction.

Edit Reaction
A reaction can hold an arbitrary number of products and reactants. Each product/reactant is defined by its stoichiometry, the metabolite, and the compartment. To select a metabolite type in the first characters and the system will present you a list of matching metabolites.


Fig: Suggestion box for meatbolites.

One reaction can belong to only one model. Moreover, it is not possible to assign a reaction to a different model as this would compromise the integrity of the reconstruction.

The systems performs a balance check that controls if there are as many elements in reactants as there are in products. As a prerequisite the InChI code has to be present for the metabolites of the reaction. In detail, it sums up all reactant elements, and compares it to the sum of all product elements. Reactions can only be edited if the latest model version has been selected. Otherwise the edit icon will be grayed out.


Fig: Unbalanced reaction.

Metabolites

The metabolite list displays all available metabolites. Editors are able to see all metabolites in the system. Visitors and unregistered users see only accepted versions of metabolites.
View/Edit Metabolites
The view page of a metabolite displays amongst other things its compartment and the consuming and producing reactions it is participating in. Please notice that only for the current selection of models those references to reactions are displayed.

Genes

The gene list displays all available genes. In addition, for each gene it links to the respective BioCyc website (uses the organism linked to the gene and the gene-name).


Fig: Example of a gene-list entry.

Gene information
A gene can be linked to multiple databases. If a UniProt identifier (see Miriam) is present, the system allows to directly download the associated protein sequence. Moreover, pressing on Display more information opens an overflow panel displaying additional information directly obtained from the UniProt webpage.

File Zone

The file zone provides the possibility to upload new files to the system and lists all available files. Currently MEMOSys differs between 3 file categories: 1) SBML files, 2) image maps (graphical representation of a model), and 3) MISC files (all other file types). Moreover, it allows users to download or delete files if the user is a member of the editor group.

Misc

Databases
The web server allows the definition an arbitrary number of links to external databases for reactions (see section Reactions), metabolites (see section Metabolites), and genes (see section Genes). Each database type has to be previously defined in the system. The databases list shows all defined external database links whereas each entry consists of a name and a URL. The reference itself can be provided either as a MIRIAM annotation or as a plain hyperlink. \textit{Note:} the application does not submit new data types to MIRIAM resources.
Organism
MEMOSys lists all organisms of models where the user has been given access rights (either explicitly in the usermanagement or by public available models). In order to avoid inconsistencies, MEMOSys lists all available organisms when an editor is creating a new model and assigning it to an organism (Model - Create new model).
Organism Edit
When editing an organism the correct BioCyc ID can be assigned. The right field is an auto-completion field that fetches all organisms from the BioCyc website. By typing in the first letters of the organism the field is populated and the correct organism can be selected. Upon selection the left field is filled with the correct ID.


Fig: Editing of an organism.

Webboard
The webboard allows each user to create new threads and post comments to existing threads. Threads are shown to every user of the web server and each thread can contain an arbitrary number of comments. They can be marked as sticky which permanently keeps them on the first page.


Fig: Webboard - shows list of threads.

Threads can be associated to entities by clicking on Create new Thread in the Threads section on the detail page of the entity. The link is then shown on the detailed page of the corresponding thread.


Fig: Thread - shows the thread and its comments. The link to the associated subsystem is shown beneath the author.

Versioning

Versioned Entities

The system automatically stores the history of the following entities:
  • Metabolite
  • Reaction
  • Model
  • Author
  • Citation
  • Gene
  • Subsystem
  • Compartment

Display History

When viewing a versioned entity the web server provides a functionality to display the history of the shown entity.


Fig: Example history of an entity.

By clicking on the History panel the versions of the entity are shown in a list. Displayed are the following properties:
  • Version number - an internally used version number of the entity
  • Comment - a user comment about the performed changes
  • Date - timepoint of the change
  • User - the user performing the modification
  • Accepted - if this version was accepted at that time
  • Differences - displayed are the differences to the last version

Quicksearch

Search

The quicksearch input field is located on the left side of the screen below the navigation panel.


Fig: Quicksearch input box beneath the menu section (on the left side of the screen).

If the checkbox next to search (labeled all) is ticked than all reactions are searched for the entered keyword. Otherwise only reactions in selected models are considered in the search.

Quicksearch allows you to search for:
  • Reactions
    • Abbreviation
    • Name
    • EC Number
    • Enzyme
    • Kegg ID
    • ORF
  • Metabolites
    • Abbreviation
    • Name
    • ChEBI ID
    • KEGG ID
    • Formula
  • Genes
    • Abbreviation
    • Name
    • EC Number
    • KEGG Map
    • Swissprot ID
    • Uniprot ID
  • Organisms
    • Name
The search term need not to be complete, quicksearch is able to find entities where only a substring of the search term matched. Currently, all miriam fields are considered in the quicksearch.

Results

The results of the search are presented in one page, whereas each group has its own panel (which can also be collapsed). Metabolite, gene, and organism results contain links to the associated reactions. All lists support pagination of results and are able to show all results in one page. Furthermore, you can adjust the number of items shown per page.


Fig: Quicksearch result for input 'AAC' (Organisms are not shown here).

Comparison

Selection

Currently, it is possible to compare two different models (also compare 2 versions of a model) with each other.


Fig: Comparison of models - selection screen.

The following entities of a model are compared:
  • Reactions are compared based on its KEGG ID. If the ID is not present the equations (reversibility, metabolites and their stoichiometry) are compared. Please notice that the comparison of equations is computationally more expensive and therefore your models should incorporate the KEGG ID for reactions.
  • Metabolites are compared based on their ChEBI ID, KEGG ID, and metabolite name.
  • Genes are compared based on their UniProt ID, SGD ID, and gene name.
Note: Loading not the latest version takes considerably more time than loading the latest version. Please be patient while the results are calculated.

Result

The comparison result is displayed using Venn diagrams which helps getting a quick overview of the equality of the compared models.

The next section allows restricting the results to selected compartments and subsystems.


Fig: Comparison of models - result screen.

Scrolling down the result page, lists are provided which display equal entities, unique entities in model 1, and unique entities in model 2. Using the tabs the user can switch between reactions, metabolites, and genes.


Fig: Comparison of models - detailed results.

If the compared models have been previously selected at the model selection page links are provided to the detail pages of the listed reactions. In addition, links to metabolites and genes are displayed regardless of the model selection.

User

User Classes

The following section explains the user classes that have been defined in this application.
Users without an account
The web server allows users that are not registered to view accepted versions of all entities. Moreover, only models that have been made publicly available are visible to this user. Users without an account are not allowed to create, update, or edit entities.
Visitor
A visitor is granted the same rights as a user without an account. Additionally, all settings that are stored in the database are persisted for this user and loaded on the next login. Moreover, visitors can be granted access to specific models (through the usermanagement system) for which always the latest version of entities is shown even if they have not been accepted.
Editor
An editor is granted the same rights as a visitor and is additionally allowed to create, update, and delete all entities of models where access rights are assigned (usermanagement system). Whenever a change is made to an entity its status changes to \textit{not accepted} and needs to be explicitly accepted by a superuser. Editors are allowed to upload files to the web server and import SBML models into the database.
Superuser
A superuser is granted the same rights as an editor, and all models are visible to the superuser. Additionally, the following actions can be performed:
  • Accept version of entities.
    Please note that the list of not accepted reactions - especially after a new model has been imported - can be very long, and it is therefore better to accept the whole model instead of individual reactions.
  • Change the public available state of a model.
  • Set a webboard thread to sticky (tick or untick checkbox in the sticky column).

User Settings

The web server stores individual settings of users permanently in the database and loads them when the user logs in again. Amongst others, the following settings are stored in the database:
  • Screen layout
  • Selection of models done at is stored.
  • Sort order of lists

Access to models

MEMOSys uses a dedicated usermanagement system to provide users/groups fine grained access to models. In order to grant certain users access to models please follow the steps documented at section Model Rights.

SBML

Import SBML

The import SBML section allows users to import a model stored in the SBML format into the database. The SBML format has to be at least Level 2 version 3. Before a file can be parsed, it needs to be uploaded to the system. By pressing the parse button the selected file is parsed and results are written into the database. A progress bar informs the user about the ongoing parsing job. After the import is finished, statistics are displayed that provide general overview information about the import process.

Export SBML

There are two different ways to export (parts of) a model into an valid SBML file.
  1. Export the complete model (see section Export Model)
  2. Export a previously selected reaction list (see section Export Reaction List)
Export the complete model
  • Model - select the model you want to export into SBML format. All reactions of this model will be included into the SBML file
  • Version - select the version of the selected model
  • Subsystems - select the subsystems that should be exported. If no subsystem has been selected all subsystems will be included into the file. The selection is filled as soon as the model and the version are specified.
  • Compartmentalization
    • Compartmentalized reactions and metabolites are completely compartmentalized as they are stored in the system.
    • Partially Decompartmentalized reactions and metabolites that are in subcompartments of the Cytosol are assigned to the Cytosol. The Extracellular compartment is untouched.
    • Fully Decompartmentalized there are no compartments - unsegregated system.
Export ReactionList
Reactions can be exported into a SBML file by pressing the export button displayed on the reaction list page. This allows the user to perform any kind of query on reactions and export the resulting list into a SBML file.
The following selections need to be performed:
  • Model select the model to export.
  • Compartmentalization

Improve Annotation

The improve annotation functionality is available to all editors and can be used to add or modify certain properties of an entity. It is accessible via a link in the SBML menu.

Select file and start

The setup page allows the user to select an uploaded SBML file, and starts parsing and generating results when the \textit{start} button is clicked.


Fig: Selection and start to improve annotation of entities.

Result

The tabs on the top allow switching between metabolites and genes. All actions performed in a tab affect only the selected entity type. Displayed are lists of entities which hold 35 items per page and support full pagination.
The following buttons are available for each entity type:
  • Set all to new - sets all properties to the loaded value of the property.
  • Set all empty properties - sets all properties that are empty (no value stored) to the loaded value (if the loaded one is not empty).
  • Save xyz - saves the current selection.


Fig: Result of annotation improvement: abbreviation and name show scenario 1; 3rd row of KEGG and ChEBI show scenario 2; 1st and 2nd row of KEGG show scenario 3.

For each displayed property the following scenarios exist:
  1. The stored value and the new loaded value are the same for the property - the value is just displayed.
  2. The stored value differs from the loaded value - two radio selections are shown where the first one shows the stored value and the second one the loaded value. According to the selection, either the stored or the loaded value will be saved for the property.
  3. No value is stored and nothing is loaded - field is empty.
Metabolites
The following properties are listed for metabolites:
  • Name
  • Abbreviation
  • Synonyms
  • Kegg Id
  • Chebi Id
  • Formula
  • InChi
Genes
The following properties are listed for genes:
  • Name
  • Abbreviation
  • EcNumber
  • Swissprot
  • Uniprot
  • Sgd
  • KeggMap