Clover Bioanalytical Software

Granada

Clover MS Data Analysis Software

User Manual

Lite Version
Go to Title ↑

For all content you can consult Full Manual Version

(Clover account required)

Regulatory and Legal notices

Copyright© 2015
All other trademarks are the sole property of their respective owners.

All Rights Reserved
Reproduction, adaptation, or translation without prior written permission is prohibited, except as allowed under the copyright laws.

Warranty
The information contained in this document is subject to change without notice. Clover Bioanalytical Software makes no warranty of any kind with regard to this material, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. Clover Bioanalytical Software is not liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance or use of this material.

Document History
Clover MS Data Analysis Software User Manual, Version 1 (August, 2022) First edition: August 2022

Limitations on Use
For Research Use Only (RUO). Not for use in diagnostic procedures.

Hyperlink Disclaimer
Clover Bioanalytical Software makes no express warranty, neither written nor oral, and is neither responsible nor liable for data or content from the linked Internet resources presented in this document.

Clover Bioanalytical Software

Edif. Centro de Empresas PTS
Av. del Conocimiento, 41
18016 Granada, Spain
Phone: +34 958 991 543

cloverbiosoft.com
clovermsdataanalysis.com
info@cloverbiosoft.com
support@cloverbiosoft.com 

1. Scope

This document has been written with the aim of helping users to handle Clover Biosoft web application. Our platform can perform the analysis of mass spectrometry and infrared spectroscopy data, and it is mainly focused on Strain Typing, Biomarker Discovery research and the use of Machine Learning algorithms.

Reminder: If you do not have access to our web application you can sign up for a free trial by filling out our form with your main information (Figure 1.1). The reading and the acceptation of our terms of use are required to complete the registration process. We will send you an email as fast as possible with a link to set a new password to complete your registration.

Figure 1.1: Sign up form

2. Login to the platform

After registration, you should receive an email with an activation/confirmation URL to enter for the first time to the platform. Type in your email address and your password into the corresponding fields (Figure 2.1). Click on the Login button or press the Enter key to authenticate. There is also an account recovery section in case you have forgotten your password (Figure 2.2). For the next logins you can use the main URL of the platform:


Figure 2.1:  Web application login form

Figure 2.2: Account recovery form


A red asterisk means that the corresponding field is mandatory.

3. Getting started

Figure 3.1 shows the general view of the web application. There you can find the following blocks of information.


  • 1 - The navigation bar: The main access to each one of the web application functionalities. It can be hidden or made visible again by pressing the top left corner icon [≡].
  • 2 - User private information:
    • Number of studies that the user has created and those that is sharing with its work colleagues.
    • Number of created and shared experiments.
    • Number of created and shared peak matrices.
    • Used storage. The usage limits its 40 gigabytes.
  • 3 - A list with all the users of the organisation, including their complete names and their email addresses.
  • 4 - Another list with the created folders and uploaded files.

Name, email and organization of the current user can be found at the top right side of the layout, next to the Logout button. The current version of the web application can also be checked in the bottom left side. This information together with the navigation bar are always visible, regardless of the selected section.


Figure 3.1: Main layout of the web application, 1. Navigation bar, 2. User private information, 3. Organization and its users, 4. Tree list view of uploaded data

4. My profile section

In this section you have access to:

  • The date you joined the platform and the complete name and email address used for the registration.
  • Password change. Click on the Change password link (Figure 4.1) and type in the old password, the new password twice (for reconfirmation) and press the Change password button to apply the changes.
  • The list of members in your organization (same structure as the Work colleagues list of the main layout).

Figure 4.1: My profile section layout. Within the green circle is the Change Password link

5. Upload files section

In this section you can manage all your data and upload new one. Currently, Clover Data Analysis software support Matrix-Assisted Laser Desorption/Ionization - Time of Fly (MALDI- TOF) Mass Spectrometry and Fourier Transform Infrared (FTIR) spectroscopy data in their most common format.

  • Management of data
    • Create new folder. You can create as many folders and subfolders as you need to manage and sort your data. The tree list view allows you to navigate across your folders (Figure 5.1). You can click on New folder in "root" (Figure 5.1a) to create a main folder where you can upload directly your files or create one or more subfolders within it by clicking the folder icon (Figure 5.1b).
    • Edit and delete folders. Any folder can be renamed (pencil icon) or deleted (bin icon) with all its content (Figure 5.1b).
  • Upload new spectra files
    1. Choose file source between MALDI-TOF MS or FTIR technologies (Figure 5.1-1).
    2. Choose file format between Open data format or Bruker format (Figure 5.1-2).

    3. This step only appears when you choose MALDI-TOF MS technology.
    4. Select the destination folder (Figure 5.1-3). You can choose between the hierar- chy of folders that you have already created, create a new folder (Figure 5.1-a) or upload the files in folders to the root. The root folder is selected by default.

    5. Is not recommended to upload directly files to the root to keep it clean. This makes the management of the data within the studies and the creation of the categories easier.
    6. Click on the Choose button (Figure 5.1c) and select the file or files to be submitted in the File Explorer, or drag them manually to the Drag and Drop area. There is a size limit of 50 MB for each uploaded file. You can also Clear the list of files to be uploaded by pressing the corresponding button (Figure 5.1-4). Click on the Submit button to complete the process and upload the selected files.

    7. Please note that time needed to complete the process will vary depending on the internet connection, the size of your files and the processing per- formed by the platform. Bruker format will be transformed to .mzml files.
Figure 5.1: Upload files section layout

Up to this document version, supported file formats are:

  • For Mass Spectrometry:
    • .txt: Standard text document with plane text.
    • .xls: Spreadsheet file with data displayed in a table format. It is the old version of the .xlsx document.
    • .xlsx: It has the same format as the .xls document but in a much smaller size.
    • .mzml: New format which aims to marry the best elements of .mzxml and other mass spectrometry documents.
    • .mzxml: XML based format for encoding mass spectrometer output files.
    • .csv: Spreadsheet file with plain text data sets separated by commas (open source).

    Please note that files/spectra from Bruker™ instruments need to be up- loaded within a .zip folder.
  • For Infrared Spectroscopy:
    • .xml: Exported by Bruker OPUS software.
    • .csv: Spreadsheet file with plain text data sets separated by commas (open source).
    • .spr: Perkin Elmer binary format exported by in Perkin Elmer IR.

6. Spectra files section

In this section, you can see your current list of files and folders in a tree list view (Figure 6.1). It also includes information about the size and the upload date of each file (Figure 6.1a), the total number of uploaded files (Figure 6.1b) and the total files selected, if any (Figure 6.1c). This section allows the files and folder management:

  • There is a checkbox on the left side of each file and folder to enable their selection. All objects can be selected by clicking the Select all button.
  • Selected objects can be removed and downloaded. Folders can also be deleted by clicking the bin icon on the right side of the row (Figure 6.1-1).
  • It is possible to create a new folder at the root level by clicking the New folder link or to create a subfolder by clicking the folder icon on the right side of the parent folder.
  • List sorting by data, name or size. The order is ascending by default, but you can change it through the icon on the right side of this functionality.
Figure 6.1: Spectra files section layout

Through this section you can also access to the File Viewer (Figure 6.1-2) by clicking on a file name, test_sample in this example. This section shows the selected spectrum in a dynamic viewer. Files from the same folder can be also shown directly (Figure 6.1d) and a 3D View is available by switching the 3D button (Figure 6.1e).

7. Data management and Organization

Data management and organization in Clover Data Analysis Software platform has a fairly similar handling to any laboratory (Figure 7.1).

Studies are the highest level of data distribution in the platform, they can contain MS or IR spectra (not both). Each study can be shared between colleagues of the same institution or with other users if you know the email they used to register into the platform. By this way, the content of the study can be consulted by all the users the study was shared with. In addition, they can create experiments, peak matrices and run any analysis inside that study. Experiments are the second level of data distribution.

Experiments are located within the studies and they may contain all or some of the data inside the study. Thus, you can have a big study for your research which contains all your data, and within that study all the exper- iments you need, each one with its own data from your study. Each experiment has its own preprocessing and it will be applied to all of its spectra.

Peak matrices can be constructed from the experiment data. They are the input data for the Machine Learning Algorithms available on CLOVER platform. The results of these algorithms trained with the peak matrices can be saved as Prediction Models and tested in our Validation Module with external samples. These Prediction Models are linked with each study (not with the experiments).


Figure 7.1: Data Management diagram

8. New study section

This section is focused on the creation of new studies, which are the highest level of organization in the management of data in CLOVER platform. These studies, and all the spectra and metadata that contain, are the ones which can be shared with users within your organization or with other users in the platform.

Each study has a name and a description (maximum string length is 1600) that you must fill. After that, you must choose between MALDI-TOF (Mass Spectrometry) or FTIR (Infrared Spectroscopy) source*. Once you have selected the files and folders that you want to include in the study, press the Submit button to create and save your study with these data.


Figure 8.1. New study section layout

*Please note that folders and files will be filtered according to your choice. Each study can only contain files from one type of source.

9. Studies section


Figure 9.1: Studies section layout

Here, you can see your list of studies. You can organise them by private or shared studies (Figure 9.1-1). You can also filter this information by only showing the studies that you have created. The study data is indicated as follows in this section:

  • The name and the description of the study. The name is also a link that will redirect you to a new layout with extensive information of the study (Figure 9.1-2).
  • The creator of the study and the creation date (Figure 9.1-3). The data source (MALDI-TOF or FTIR), the number of users with access to the study and the number of files that are part of the study are also displayed in this box. Click on the number of files to see detailed information about them. You can organise the information as a flat list or as a tree as well as download all or some of the files (Figure 9.2).
  • It is possible to create a new study within this layout too by clicking on Create Study (Figure 9.1-4). This action will take you to the Studies section.

    If no studies are created yet, this section displays just the option to create the first study.

Figure 9.2: Detailed information about the files of a study

9.1. Detailed study section

Figure 9.3: Detailed study layout with an experiment subsection that has various experiments with peak matrices

As mentioned before, if you click on the name of a study, a new layout will be displayed with all the study information. This is the main layout of the selected study, which contains all its information, data and options.


  1. General information of the study: name, description, creation date and owner of the study (Figure 9.3-1).
  2. Share and collaborative options (Figure 9.3-2).
    • The data source (MALDI-TOF in this example).
    • Team (x users). The number of users with access to the study. If you click there, the platform will display a new layout with the user information, i.e. email, name and organization (Figure 9.4).
    • Figure 9.4: Detailed information about the users that are following a study
    • Add Users. A new layout appears when clicking here. You can share the study with the coworkers of your organization just by selecting the box to the left of the user name or using the Select All button (Figure 9.5a). If the user you want to share the study with is out of your organization, you must select the With other option and write the registered email address of the user (Figure 9.5b).
    • Figure 9.5: Share Study form. a) Within your organization layout. b) With others users layout
    • Delete Study. This option deletes all study content and configuration, but files will be still uploaded in their respective folder in the platform. The platform will advise you if you are sure to do it or if there is any conflict that prevent you to do it.

      Please note that only studies created by yourself are susceptible to be deleted.
  3. Study content and metadata (Figure 9.3-3). These options will take you to their subsec- tions within the study (Figure 9.6). The Experiments subsection is displayed as preset. The list of experiments of a new study is empty at the beginning. You have to click on the here link to create a new one. This part will be explained in Experiment section later.
  4. Figure 9.6: Study tabs detailed view
    • Experiments. Tab with the list of experiments that are part of the study and their peak matrices. You can order the experiments by date or by name, and there is also a link for creating a new experiment.
      If you click on the left arrow next to the experiment name (Figure 9.3a), a dropdown will appear with the peak matrices created within that experiment and their main parameters. Detailed information about these parameters can be consulted placing the cursor above it (Figure 9.3-f). Furthermore, you can download, delete or mark as favourite any of your peak matrices (Figure 9.3c).
      You can also go to the experiment subsection to amplify the information of the selected experiment by clicking on its name (Figure 9.3b). In the other hand, new peak matrices within the selected experiment or new experiments can be created directly in this subsection by clicking on New Peak Matrix (Figure 9.3-d) and Create Experiment (Figure 9.3-e) buttons respectively. Both processes will be described in Peak Matrix generation section and Experiment subsection as com- mented above.
    • Files. Tab with the number of files included on the study. Click on it if you want to see more information about those files showed in the tree list view, or bring previously uploaded files to the study (Figure 9.7a) as well as remove folders or files from it (Figure 9.7c).
    • Figure 9.7: Files Subsection layout within Study View

      Please note that only studies created by yourself are susceptible to be deleted.
    • Categories. Tab with the tools needed to create and manage the categories of the study. This tab will also be explained with more details in Categories subsection.
    • Prediction Models. Tab with all Prediction models saved at the study. The results of algorithms can be saved as these prediction models. They can be validated and used for identifying new samples. This tab will be further explained in Prediction Model subsection.



9.2. New Experiment setp

After the study is created, you can create as many experiments within it as you need. For this, click on Next Experiment button in the Study View layout (Figure 9.3e), this action takes you to New Experiment step. This layout is pretty similar to the New study one (Figure 9.8). The only new thing here is the Experiment metadata. You can write whatever information you want here. Type in a key name and a value for it and press the Add button to add the pair to the experiment. Press the Save button to save your experiment.

Figure 9.8: New experiment step layout

Please note that only the folders and files that are included in the study will appear here.

Once you have created the experiment, the platform will take you to the experiment layout (Figure 9.9). This experiment will be empty (with no peak matrices and no preprocessing). You must choose then between preprocess the spectra attached to the experiment, or create a new peak matrix. The usual procedure is to apply a preprocessing before creating any peak matrix, thereby, all peak matrices generated after this will be equally preprocessed and their replicates (if applicable) managed. Both processes will be described in detail in Preprocess data process and Peak Matrix Generation process.

Figure 9.9: Empty experiment layout

Please note that only studies created by yourself are susceptible to be deleted.

The third option that you can do without creating a peak matrix or applying any preprocessing is a Biomarker Analysis/Technical analysis. Although a preprocessing with noise reduction applied is always recommended, you can run this analysis with raw data. Either checking how preprocessing affect to your data, making a reproducibility analaysis in a particular point of your replicates or if the preprocessing is already applied by other software. These analysis will be explained in details in Biomarker/Technical Analysis.



9.3. Detailed experiment subsection

If you click on the experiment name in the Detailed study section (Figure 9.3b) its specific experiment section layout displays with all the related information about the selected experiment.

Figure 9.10: Detailed experiment layout with peak matrices

This subsection contains all the experiment information:

  1. The navigation path, which shows the experiment name and the name of the study it belongs to.
  2. The information of the experiment: name, a short description, the creator name, the creation date and the number of the files that are part of it. All this information can be edited and modified by clicking on the pencil shown when the cursor is over the information that you want to modify. You can also click on files if you want to see more information about those files in a similar way than files in studies (Figure 9.2).
  3. Dropdowns. Section with additional information and experiment tools shown when you click on the arrow next to the section name.
    1. The Preprocessing parameters of the experiment. These parameters are applied to all peak matrices created within this experiment.

      Preprocessing can not be edited but it can be deleted and done again. This option is only enabled when the experiment has no peak matrices.
    2. Spectra Visualizer. This tool allows you to show all the spectra within the study. If the preprocessing is applied, the spectra shown will be the preprocessed ones. If, on the other hand, the preprocessing was skipped or is not done, the spectra shown will be with no preprocessing. You can check detailed information about CLOVER ́s viewer in its corresponding annex.
    3. Metadata. The key and value pairs that make up the experiment metadata (Figure 9.11). You can add all the information you need by clicking the button Add+ inside the section.
    Figure 9.11: Metada dropdown information
  4. Peak matrices. This section displays all the peak matrices created inside the experi- ment. The option Show only my peak matrices only appears when the experiment is shared and other users have created peak matrices. The information displayed here is the same than the one shown in the detailed study subsection (Figure 9.3f). Inside this section you can find the New Peak Matrix button to start the process of building a new one. You can also order them by date or name in ascending or descending order. You can click in the section name of a specific peak matrix to go to its detailed peak matrix layout.
  5. Additional options. This option is displayed when clicking on the 3 points icon at the right upper corner. You can add new files to this experiment, customize samples by colour or delete the experiment form the database (reconfirmation modal appears for this one). If no peak matrices created, an additional option appears here (Remove files from Experiment) which allows you to remove files from the selected experiment, but not from the study.


    The files to be added to the experiment have to be included in the study before you can do this action. Also, this files will be preprocessed with the same param- eters from the preprocessing if applied.

    Biomarker Analysis can be started from here too. This analysis will be explained in its own section. Unlike algorithms from CLOVER platform, Biomarker Analysis runs by using the experiment samples as an input data and does not need any peak matrix.

9.3.1. Customize samples layout

If you click on the Customize Samples link (Figure 9.10-5), you will be redirected to a new layout which shows all samples within the experiment selected (Figure 9.12).

Figure 9.12: Customize samples layout

Here, you can change the colors used to render each experiment sample spectrum. Select the samples that you want to color (you can show them in a single list or separated by file or folder). Pick the color that you want from the color palette shown after clicking on Pick color button to apply and do not forget to apply the changes. Color can be selected by HEX, RGB or HSL color spaces. Samples will be displayed with the color selected in plots, charts and the spectra viewer.

You have other alternatives too. They will both apply the corresponding colors and save the changes automatically:

  • Apply the default color to all the samples. It sets the default dark blue color (0080c5) to all samples.
  • Apply sequential colors. Each sample has a different color.
  • Apply the owner’s color. Available only when you are following the experiment. Each sample has the current color that the owner has for it.

9.3.2. Peak matrix detailed layout

This layout is accessible by clicking on the name or in the section of a specific peak matrix from Study or Experiment sections (Figures 9.3f and 9.10-4). A specific layout displays with all the specific information about the selected peak matrix (Figure 9.3.2).

Figure 9.13: Peak matrix layout

The information shown in Figure 9.13 can be divided into:

  1. Peak matrix path and information related to its name, a short description and the user who created the peak matrix and the date.
  2. Collapsable areas with specific information and tools. You can display the content by clicking on the box header. The first one has all parameters information, divided into Preprocessing and Peak matrix parameters (Figure 9.14).
  3. Figure 9.14: Parameters applied

    From Peak Matrix Table you can show and download as .csv format the peak matrix table. Each row is a m/z value and each column is a sample with an intensity for that m/z value (Figure 9.15). You can filter by an specific value in each column as well as filter by mass range by scrolling the horizontal top green bar. The last two collapsable areas are the Heatmap and Plot of the peak matrix, both tools will be described in their respective annex sections.

  4. A direct access to Classification algorithms and PCA analysis. You can directly go to the algorithms views or obtain the PCA results from here.
  5. Figure 9.15: Peak Matrix Table

9.4. Categories subsection

In this section you can create, edit and delete categories from the selected study. A category is formed by a label, a description and a color. These categories are the way CLOVER platform labels the samples within the studies. One category includes one or several samples, and each sample could be included in more than one category at the same time Categories will be used in every analysis run within the study.

  • In supervised algorithms you can choose between the categories created in this section for further classification or discrimination.
  • In unsupervised algorithms this categories will help you to see more easily the distribu- tion of your data by coloring the samples.
  • In Biomarker Analysis the categories will be used to analyse and search potential peaks that can be used to identify or discriminate one category from other.
  • In Reproducibility Analysis you can compare the metrics and statistics between these categories.
  • In Validation module this categories will be used for determining the accuracy and the other metrics of the prediction models.


Due to the importance of the categories and its feature as sample labelling, it is recom- mended to read this section carefully.

You can access to this section from Study layout by clicking on Categories subsection button (Figure 9.16). You can create as many categories as you need in the study by clicking on the + Create Category button.

Figure 9.16: Empty Categories subsection layout

A new layout displays for creating a category with two main options:

  • Create categories from folder. (Figure 9.17a) Automatically create a category using the samples included in a specific folder. The platform shows some options to name the category from the name of the folder or its path. You can also create one category per folder at the same time by choosing multiple folders or by clicking on Select All button. Categories created by this way can also be edited manually once they are saved by adding or deleting samples or by changing their names or colors.
  • Create empty category. (Figure 9.17b) You can create an empty category and add samples to it later. Here you have to name the category, choose any color for it and, optionally, write a short description.


One sample can be in more than one category. In the example of Figure 9.17, samples from strain 2 folder can be within Strain 2 category and within Subspecies 2 category.
Figure 9.17: Category creation step. a) Create from folders automatically, b) create empty cat- egory to select samples manually

If you choose Create Empty Category option, you have to add manually all samples you want to include in that category. For that, you need to go to the specific category layout (Figure 9.18) by clicking on it (Figure 9.19). In this layout you can see all samples included in the category sorted by name or by folder as well as delete this category or edit samples. Edit Samples button redirects you to a layout where you can add or remove any samples from the category.

Figure 9.18: Specific category subsection layout


Remember, only samples included in the study will be displayed in this layout. All cat- egory changes or configuration will be applied to the entire study. A good distribution and well organized samples in folders will help you to configure the categories in a faster and easier way.
Figure 9.19: Category subsection with categories. The user can still create categories from Create Category button (green arrow).


Categories created by other users within a shared study only can be edited and deleted by the owner of the study or the creator of those categories.

9.5. Prediction Model Subsection

In this subsection all Prediction Models saved are shown (Figure 9.20). You can switch the filter from All to an specific algorithm to display only that type of algorithm as well as enable the Show only prediction models created by myself to show only your prediction models. In each prediction model box you can find the title of the prediction model, a short descrip- tion and the categories used to run the analysis. In the right side of each box you can see the owner of the prediction model, the algorithm used and the peak matrix used as an input data for the analysis, as well a bin icon to delete the prediction model if you are the model owner.

Figure 9.20: Prediction Models subsection. Each box is a Prediction Model from an algorithm previously saved by the user.

A Prediction Model is the result of a supervised algorithm. If you click on a box of a specific prediction model the platform redirects you to the Prediction Model layout (Figure 9.21). Here you can directly reproduce the analysis with the same peak matrix, categories and hyperpa- rameters (if applicable) used when it was saved. You can only change the folds of k-fold test (10-fold by default). The prediction base layout gives you all the graphs and tools needed to interpret the results.

From this layout, you can Validate the model by clicking on the button at the upper right corner. This action redirects you to the second step of Validation Module process since the first step is about choosing the prediction model.

Figure 9.21: Prediction Model subsection layout of an specific analysis.
Made with Pingendo Free  Pingendo logo