For all content you can consult Full Manual Version
(Clover account required)
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.
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
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.
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 2.2: Account recovery form
A red asterisk means that the corresponding field is mandatory.
Figure 3.1 shows the general view of the web application. There you can find the following blocks of information.
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.
In this section you have access to:
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.
Up to this document version, supported file formats are:
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:
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).
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).
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.
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:
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.
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.
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.
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.
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.
This subsection contains all the experiment information:
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.
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).
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:
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).
The information shown in Figure 9.13 can be divided into:
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.
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.
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.
A new layout displays for creating a category with two main options:
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.
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.
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.