Aryballe Suite Overview

Aryballe Suite is the proprietary software provided with the NeOse Advance used in the acquisition and analysis of odor data.

Computer specifications

The user will need a computer operating under the following minimum requirements:

• Microsoft Windows 10 Professional (64-bits), build 1909 or higher
• 8 GB RAM (16 GB recommended)
• CPU Core i5 Gen 8 and after (> 2017) 1.6 GHz
• Internet connection to download the Aryballe Suite software
• Chrome or Mozilla Firefox installed
• USB ports: 1 SuperSpeed USB

Account Activation

Before you install the Aryballe Suite for the first time, make sure that you have activated your Okta account. 

1. Click on the "Activate Okta Account" button in the email that was sent to you by Okta. You will be redirected to the Okta registration page. 

   
   

   
   

2. Once on the Okta activation page, create your password and security questions. Click on the "Create My Account" button to finalize your account creation.

This procedure will need to be repeated for each new user account created in your organization. If you have any questions about account set up, please contact support@aryballe.com.

Aryballe Suite installation procedure

1. Download the Aryballe Suite Setup program using the link provided via email from the Aryballe team.
   
   
2. Double click on the .exe installer and follow the wizard instructions.
   
   
3. IMPORTANT: You must uninstall any previous version of the Aryballe Suite software before you proceed with the installation. If an older version is installed on your computer, you will not be able to install the new software.
   

   
   

4. Upon software installation, a setup window will appear. Click on the “Next” button.
   
   
5.  A License Agreement window will appear. After reading the agreement, check "I accept the agreement" and click on the "Next" Button.
   
   
6. The setup window lists the components that will be installed. Click on the “Install” button.
   
   
7. The installation displays the following window once complete:
   
   
   
8. The Aryballe Suite Software will be implemented in your computer's list of applications. By clicking on the Windows icon, you will be able to open the application interface. "Aryballe Suite" allows you to both launch experiments with NeOse Advance and analyze your data from the cloud-based TSDB (Time Series Database). "Aryballe Analysis (local only)" lets you analyze data in your local database.

   
   

9. Open the "Aryballe Suite" application. A login window will open. Use your email as username and the password you have created on the Okta platform. 
   
   
10. Once you log in, the application homepage will open. The software bundle includes two modules: the Acquisition Module to run acquisitions using protocols predefined by Aryballe, and the Analysis Module to analyze data using HTML reports.
    
    
    Note: It is not possible to run two instances of the application. If you try to open the application when it is already running, a warning pop-up window will open.
      

Acquisition Module Overview

The Acquisition module executes protocols provided with the NeOse Advance. While pre-defined, the protocols can be customized to adjust experimental conditions.

The home page of the Acquisition module is organized in 3 parts:

• Activity Sequence & User action: allows users to visualize the different acquisition steps and start a cleanup protocol.
• List of devices: list of the devices that can be used. A warm-up button is available to initiate the instrument warm-up protocol which allows the device to reach a stable temperature before measurements. 
• List of protocols: list of the different experimental protocols that can be applied to samples (Odor Acquisition, Calibration & Check according to the accessories associated). A short description of the protocol is provided upon hover.

Live Data View and Annotator

The Live Data View allows users to see data measured by sensors during a run in real time, while the Annotator Tool allows users to add contextual information to the captured data during the experiment. These two tools can be accessed either within the Aryballe Suite application or via in an internet browser—Aryballe supports Chrome and Mozilla Firefox.

Live Data View

The Live Data View shows the visualization of the data captured in real time during acquisition. This versatile tool allows the user to select sensor data to display depending on the use case of interest.

When the Acquisition module is executing a protocol, click on the “Live Data View” button in the top right of the summary screen.

Then, click on the “Live Data View” button in the bottom right of the following window.  

The Live Data View window will open in another tab of the application and displays visual feedback of the device status.

It is also possible to open the Live Data View in a web browser window by clicking on the "Live Data View" button from your Windows menu.

Note that the web browser window will not open unless a run has been started.

The Live Data View is composed of two horizontal panels to simultaneously visualize two types of information.

The top banner displays the name of the device used.

On the left side of the window, users can select the parameters to display, while the right side shows sensor output values and device status information in real-time (with the last value recorded). The colors between the numerical values in the right console and in the diagram are linked.

The top and bottom diagrams are completely independent—any of the available sensors can be selected and magnified for more detail. The vertical red line indicates the point in time across both graphs. All diagrams show only the last 2 minutes by default when no magnification is applied.

The following parameters can be displayed:

• Events: inputs based on annotations entered by the user in the Annotator.
• Odor Signature: a normalized odor signature based on Aryballe’s Core Sensor response is generated at a frequency determined in the protocol (generally every 2 minutes).
• Delta Odor Intensity: the change of odor intensity between the sample measure and the baseline measure is presented as recorded by Aryballe’s Core Sensor.
• Odor Intensity: the value of odor intensity as recorded by Aryballe’s Core Sensor.
• Flow Temperature 1/Humidity 1: displays the value of temperature [°C] and humidity [%] of the sample from the HIH humidity sensor placed after the Aryballe Core Sensor in the fluidic line.

Note: when a sampler is used, the number of the opened inlet is displayed in the event window following the rule:

• 0 is baseline
• 1 is inlet N°1
• 2 is inlet N°2
• Etc.

When an Amplifier is used some parameters are added :

• Amplifier temperature: real-time temperature of the adsorbent. 
• Amplifier power: tension injected for the heating of the trap (normalized value).
• Amplifier setpoint: targeted temperature value (°C).

Annotator Tool

The Annotator Tool allows the operator to add contextual information to a measurement session. Custom events, such as additional information about specific experiments or samples, can be provided by the user in the Custom event annotation field (e.g. Sample name, Comment, Description).

1. When the Acquisition module is executing an experimental plan, click on the “Annotator” button.
2. The annotator window will be opened in a web browser. Click on the “START NEW” button of the Annotator window.
   
3. A popup window will open so that additional information can be specified. Type in the desired information and click on the “ADD” button.
   
4. Annotations will be merged with records as tags and will appear as an Event in the Live Data View interface.
   
5. To stop the event, click on the “STOP” button.
   All annotations that are overlapping a record are indicated as a tag like: $annotation: “custom event” and can be easily seen in the Analysis module browser.

Experiment Analysis Tool Overview

The Experiment Analysis tool enables visualization and statistical analysis of data captured by the solution. The tool works independently from the Acquisition module and can be used without connection to the NeOse Advance device.

The Experiment Analysis tool can be launched from the Aryballe Suite interface in a dedicated tab.

If your installed package includes the TSDB, you will work with the TSDB Experiment Analysis tool interface. For now, the TSDB is not available with the use of an Amplifier.

The homepage of the Experiment Analysis tool is organized according to the different pre-defined protocols:

• Visualization: display of odor signatures and statistical analysis
• Detection: define the Limit of Detection (LOD) of an analyte in a diluent
• Recognition: identify a sample based on learning models
• Humidity Calibration Check: validate the quality of the Humidity Calibration run
• Quality Check: validate the device performance
• Cleanliness Check: validate system cleanliness

For the Amplifier, a dedicated report is available. 

• Amplifier: display the Amplifier report

The Experiment Analysis tool can also be launched from your Windows menu by clicking on the "Aryballe Analysis (local only)" button. 
 

You will access to the local Experiment Analysis tool interface.

The experiments are organized according to runs, allowing for the selection of a particular experiment and generation of pre-formatted HTML reports. The reports include the entire set of data from Aryballe’s Core Sensor and temperature and humidity sensors.

Clustering Quality Score

CQS is an emerging metric in digital olfaction which provides the best total picture of the sensor’s ability to discriminate specific odors.

Historically, odor sensing performance was difficult to define and evaluate because there was no pre-existing reference for the absolute and widely accepted representation of odors. The adapted way of dealing with this has been to relatively compare molecules to molecules, or odors to odors.

If we assume an odor has no variability, and the sensor to measure this odor is infinitely stable, the measure of this specific stable odor would always generate a signature as an identical point in the N-dimensional space of the sensor. All the recorded odors would be like punctual stars in a sky, separated individually, as below.

As reality is often variable or non-perfect, either due to the protocol of experiment, the odor preparation, or the measurement device itself, these “stars” become fuzzy spots when repeating for a short-term period what should be the same odor measurement. This increases the probability of overlap between odors and increases confusion, hence reducing the resolution, i.e. capability of separating odors. By representing this phenomenon with a numeric value—the Clustering Quality Score—we are able to quantify the degree of separability for these odor data points. The CQS is depending on the distance between the measures of a same odor and the distance with the measures of the N other odors.

CQS allows us to “measure” the performance of a complete experiment, on a continuous scale (from -100% / very bad, to +100% / perfect). There are two different levels of CQS which are relevant to monitor:

• The Global CQS, the score representing the quality of the global separation, i.e. the complete partition. On a set of N different clustered odors, it is represented only by 1 CQS score.
• The Individual CQS, the score representing the quality of each individual cluster (which contains what is supposed to be identical measurements) among the other clusters. On a set of N different clustered odors, it is N different numbers.

The information conveyed by CQS can also be used to measure the repeatability and reproducibility of a technology by looking at records across multiple days for the same device (Repeatability) or by analyzing the score for multiple devices on the same day (Reproducibility).

Connect to Remote Desktop

Prerequisites to connect from user’s computer to remote computer.

On the remote computer’s side:

• It is on and running.
• The Aryballe Suite has been installed on the remote computer.
• All devices and accessories required for the user’s experiments are connected to the remote computer.

On user’s side:

• They have rights to connect to it and know its IP address.
• User’s computer is connected to the same local network as the remote computer, through a VPN if necessary.

1. From user’s computer, open Start menu and search for “Remote Desktop Connection”.
   
   
   
2. Open the application and fill in the IP address:
   
   
   
3. Click on “Show Options” and fill in User name if necessary. The user can also customize the display configuration in the “Display” tab.
   
   
4. Click on connect. Once the connection is made, the user can launch Aryballe Suite as usual.
   
   
5. Whenever the user needs to, he can disconnect from the computer by clicking on the “X” button:
   
   The remote computer will still run the experiment and can be connected to by any users at anytime.

Preparing your experiment

Before starting any experiment, we strongly recommend performing a full system validation to ensure that the system is fully functional and provide traceability for future reference.

Fluidic installation

1. NeOse Advance without sampler: environment measurements or sample measurements in a vial will require manual operation from the user.
The set-up will be organized as below, with a PTFE filter on the Baseline inlets of NeOse Advance. 

2. NeOse Advance with sampler: with the use of HeptaValve Mini sampler, automated measurement of up to 7 samples (in vials). The set-up will be organized as below, with a PTFE filter on the Baseline inlets of NeOse Advance and HeptaValve Mini.

3.  NeOse Advance with Amplifier: environment measurements or sample measurements in a vial will require manual operation from the user.
The set-up will be organized as below, with a PTFE filter on the Baseline inlets of NeOse Advance and of the Amplifier. 

4.  NeOse Advance with Amplifier with sampler: with the use of HeptaValve Mini sampler, automated measurement of up to 7 samples (in vials) after amplification. The set-up will be organized as below, with a PTFE filter on the Baseline inlets of NeOse Advance, HeptaValve Mini and the Amplifier.

Device Start Up

Before running any experiment, make sure that a non-sterile PTFE Luer filter is added on NeOse Advance Baseline inlet or any Baseline inlet from bundle accessories such as Heptavalve Mini or Amplifier. When drawing in ambient air, always use a filter to protect the sensor from external particles and pollutions. The PTFE filter should remain on the baseline inlet at all times while the device is operating to avoid premature aging of the sensor.

Allow the device to warm up for at least 30 minutes:

1. Connect the NeOse Advance to the computer Super Speed USB port using the USB cable provided, with the blue sticker facing upwards. Do not plug the NeOse Advance to a USB hub since it may not receive the power required to operate.

2. Open Aryballe Suite by clicking on the Aryballe Suite App from the Windows menu.

3. Open the Acquisition module by clicking on the "Start a new acquisition" button.

4. Select the device on the left part of the screen and click on the "Warm up" button.

If your NeOse Advance device does not appear in the device list, try unplugging and replugging it.

5. A pop-up window will open to make sure you warm up your NeOse Advance for at least 30 minutes before starting an experiment. Click on the "Start warm up" button.

6. A second pop-up window will open with the 30 minute countdown.

The warm up period can be skipped by clicking on the "Close" button. Please note that Aryballe cannot guarantee the quality of records that were obtained after bypassing this warm up time. 

By clicking on the "Close" button, you will return to the Acquisition page.

Allow the Amplifier to warm up for at least 30 minutes:

1. Connect the Amplifier to the computer USB port using the USB cable provided and connect the alimentation cable provided to a wall plug.

2. Start the Amplifier warm up by clicking on the "Amplifier Warming-up" button from the Windows menu.

3. A console window will open to follow the steps of the warm up.

4.When warm up is complete, hit the Return key to close the window.

System Checks 

Validating System Cleanliness 

This validation step is highly recommended to ensure the system is properly cleaned prior to making any measurement. It helps prevent system contamination and ensures data quality. Aryballe does not guarantee results from experiments performed on dirty systems. During this procedure, make sure that no vials are connected to inlets.

Cleaning Protocol of the sampler

The cleaning protocol is performed using an external pump box, without connecting the NeOse Advance.

1. Click on the "Aryballe Suite" button in the Windows menu.
   
   
   
2. Click on the Experiment icon in the left lateral bar. 

   

3. Connect the HeptaValve Mini to the computer using the USB cable provided. Make sure that no vials are connected to sample inlets. Note that several HeptaValve Mini systems can run on the same computer at the same time for this cleaning step.
   
   
4. In the Acquisition module, click on the “Clean samplers” button.
   
   
   
5. A pop-up window will open that details the next steps.When the HeptaValve Mini is connected, the “Discovering samplers” step will launch automatically.
   
   
   
   
6. If no HeptaValve Mini is connected to the computer, the pop-up window will inform you that no sampler was found.
   
   
   
7. Next, connect the USB cable of the external pump box to the computer. Connect the pump box to the HeptaValve Mini by pushing the PTFE tubing pre-assembled with a black Luer connector into the pump box inlet. Press the black button to turn on the pump box.
   
   
   Click on the “I have connected the pump. Start now!” button.
   
   
   
8. The cleaning protocol starts and each step can be followed up in the window.
   
   During the cleaning process, each inlet is alternatively switched on and off and rinsed with air for 6 seconds until the end of the specified cleaning duration.
   
   
9. Click the “Stop cleaning” button to stop the cleaning protocol before the end of the cleaning duration.
   
   
10. Disconnect the PTFE tubing pre-assembled with a black Luer connector of the external pump.
    
    
11. Turn the pump box off.

Check Cleanliness of the sampler

1. Turn on and connect the NeOse Advance to the HeptaValve Mini using the PTFE tubing pre-assembled with a black Luer connector.
   
   
2. Open the Acquisition module by clicking on "Start a new acquisition" button.
   
   
   
3. Select the device on the left part of the screen.
   
   
   
4. Select the protocol “Cleanliness check of external Sampler” on the right part of the screen, and click on the "Next" button.
   
   
5. A window will open allowing to customize some of the parameters of the protocol. Click on the "Run" button to start the protocol.
   
   
6. The steps for the Cleanliness check can be followed in the window that opens. The process will begin immediately with one measurement performed for each inlet.
   
   It is possible to follow the odor signal in Live Data View or to generate a cleanup report when the Cleanup check process is completed.
   
   
7. Click on the “Live Sensor View” button to assess the system cleanliness in real time. The Live Data View window will open in the dedicated tab of your interface, or you can open the Live Data View in your internet browser (as described in the Live Data View and Annotator paragraph).
   
   
   • If the odor intensity signal presents a peak upon valve switching, then the system is dirty and another cleaning session needs to be done.
     
     
     
   • Once the system is clean, the odor intensity signal should be stable (presented as minor fluctuations with very small intensity value, typically < 0.2 as shown on the Y axis of the chart).
     
     
     
8. At the end of the cycle, the cleanup check procedure is completed. The validation of the end of the experiment appears on the top left of the summary screen.
   
   
   
9. Click on the "Close" button to return to the Selection screen.
   
   
10. Result of the Cleanup procedure can be visualized using the Experiment Analysis tool.

Check Cleanliness of the Amplifier

This validation step is highly recommended to ensure the system is properly cleaned prior to taking any measurement. It helps prevent system contamination and ensures data quality. Aryballe does not guarantee results from experiments performed on dirty systems.

During this procedure, make sure that no vials are connected to any inlet. This protocol allows to make sure that there is no sample pollution remaining on any fluidic line or trapped on the adsorbent from a previous run of experiment. Since the Tenax TA sorbent is in contact with air, even without air flow, some VOCs may bind to Tenax between experiments, through passive diffusion.

1. Connect your NeOse Advance to the Amplifier as described in the Amplifier section of this User Manual.
   
   
2. Connect the NeOse Advance and accessories to your computer.
   
   
3. Open the Acquisition module.
   
   
4. Select the device and the protocol “Cleanliness check (with Amplifier)” and click on “Next” button.
   
   
   
5. A window will open, click on “Run” to start the protocol.
   
   
   
6. The steps for the Cleanliness check can be followed in the window that opens. The process will begin immediately with a baseline and a thermo-desorption step without concentration.
   
   
   
7. Click on the “Live Data View” button to assess the system cleanliness in real time.
   
   Once the system is clean, the odor intensity signal should be stable (presented as minor fluctuations with very small intensity value, typically < 0.2 as shown on the Y axis of the chart).
   
   
   If the odor intensity signal presents a peak upon valve switching, then the system is dirty and another cleaning session needs to be done.
   
   
   
8. At the end of the cycle, the cleanup check procedure is completed. Close the window by clicking on the "Close” button.
   
   
   
9. Result of the Cleanup procedure can be visualized using the Experiment Analysis tool.

Sensor Performance Evaluation - Quality Check

Quality Check is performed to assess the performance of the system, which can decrease over time or after pollution events. Three golden reference samples are provided by Aryballe, which should be well detected and discriminated by the NeOse Advance device.

1. Turn on the NeOse Advance and make sure to allow 30 minutes of warm up.
   
   
2. Follow previous instructions about fluidic installation (NeOse Advance associated to the Heptavalve Mini).
   
   
3. Connect the 3 golden reference samples to 3 sample ports of the HeptaValve Mini:
   • Place the extremity of PEEK tubing from the HeptaValve Mini into the reference molecule vials, going through the septum.
   • Place a vent system, such as a flat needle provided, in the septum of each vial.
   • Samples connected to inlets 1, 2 and 3 are measured alternatively so the headspace of each vial can be replenished between 2 measurements.
     
     
4. Open the Acquisition module by clicking on "Start a new acquisition" button.
   
   
5. Select the connected device.
   
   
6. Select the protocol to run “Quality check” and click on the "Next" button.
   
   
7. In the following window, run parameters can be customized.
   When checking sensor performance with the provided reference molecules GOLD4, GOLD6, GOLD7, Aryballe assigns by default GOLD4 to inlet 1, GOLD6 to inlet 2 and GOLD7 to inlet 3 to enable future performance comparisons. The name of the experimental plan can be customized to make future retrieval easier. Per default, an example is done with A, B and C.
   
   
   

   Note

   The number of cycles must be defined for the run—a minimum of 3 cycles is recommended, but 10 cycles is the default value for the Quality Check protocol.

   
   
8. Custom tags can be assigned to all records of a run to provide information on experimental setup and conditions that are being applied to the run. Specific tags can be added independently for each inlet.

   
   

9. Steps of the Quality check protocol can be tracked in the window that opens. The process begins immediately with one measurement performed for each inlet.

   Click on the “Live Data View” icon to monitor acquisition in real time. The Live Data View window will open in the dedicated tab of your interface, or can be opened in your internet browser (as described in the Live Data View and Annotator sections).

   
   
   
10. Once the run is finished, the confirmation of the run completion appears in the main window.
    
    
    
11. Click on the "Close" button to return to the Acquisition screen.
    
    
12. When the protocol is completed, disconnect the vials from the inlets by gently pulling the tubing out of each vial and removing the vent system. Disconnect the NeOse Advance from the HeptaValve Mini by unscrewing the Luer lock on the Analyte inlet of the NeOse Advance while turning the connector of the PTFE tubing in the opposite direction. Make sure that the NeOse Advance Luer lock remain tightly screwed and has not become loose in the process so as to prevent any leakage along the fluidic line.
    
    
13. Start a clean-up process as described in the Cleaning protocol section.
    
    
14. The result of the sensor performance evaluation (Quality Check) can be visualized with the Experiment Analysis tool.

To confirm the NeOse Advance is at its optimal performance level, the Clustering Quality Score (CQS) should be considered. If the device is not able to discriminate the golden reference molecules (i.e. CQS < 50%), Aryballe does not guarantee results of future experiments. If this is the case, please contact your Aryballe FAE or support@aryballe.com.

System calibration

Humidity Calibration

Similar to other sensors, Aryballe’s sensor is sensitive to humidity. Humidity Calibration evaluates the reaction of the NeOse Advance to different humidity levels and applies a humidity correction to each biosensor to ultimately minimize the impact of humidity on the odor signal. A minimum of 5 different humidity levels is mandatory, and the sample humidity level should be included in the range of the humidity calibration.

We recommend performing a Humidity Calibration every week.

Two setups can be used for humidity calibration: (a) a proportional valve connected to a vial of water, allowing to progressively change the humidity percentage to be assessed by the NeOse Advance, or (b) an HeptaValve Mini, using vials having different humidity levels.

1. Connect your system of calibration:
   • (a) Directly to the sample inlet of the NeOse Advance: a dilutor (between ambient air and a vial with water) is connected to the sample inlet and the user uses a tap to manually change the level of humidity taken in.
     The baseline inlet of the NeOse Advance is used to measure air for all measurements while the sample inlet is used to measure the different humidity levels.
     
   • Or (b) to the HeptaValve Mini: the user prepares at least 5 vials with different levels of humidity and connects each vial including a vent system, one after the other, to inlet 1, according to the instructions provided in the Acquisition module interface.
     
     The baseline inlet of the HeptaValve Mini is used to measure air for all measurements while the sample inlet 1 of the HeptaValve Mini is used to measure the different humidity levels.
     
     
2. Open the Acquisition module by clicking on "Start a new acquisition" button.
   
   
   
3. Select the device.
   
   
4. Select the protocol “Humidity calibration”.
   If using a dilutor system, select the protocol without sampler.
   If working with vials with different levels of humidity, select the protocol with sampler.
   
   
5. A window to define the run parameters will open to customize the acquisition duration. The default duration is 10 seconds.
   As with other protocols in the Runner, custom tags can be added to all the records to further detail the experimental setup and conditions being applied to the run. Click on the "Run" button to start the Humidity Calibration.
   
   
6. The Humidity Calibration protocol will start and each step can be followed in the new window opened. The default setting for this protocol is 100 cycles but the minimal number of cycles necessary is 5 corresponding to 5 different levels of humidity. Data can be viewed in real time by clicking on the “Live data view” icon.
   
   
   
7. Once the experiment starts, a pop-up window opens before each measurement to ensure that the user has changed the humidity concentration for the next measurement, either by turning the tap of the dilutor system or by placing a different vial under the inlet 1 of the HeptaValve Mini.
   
   
   
8. When the sample is ready, click on the “Yes” button.
   NOTE: The protocol will not proceed to the next measurement unless the user clicks on the “Yes” button.
   
   
9. Once you click on the "Yes" button, a pop-up window will confirm the action and the protocol will proceed.
   
   
10. Measurement of the baseline starts, followed by measurement of the sample.
    
    
    
11. When sufficient measurement of various humidity levels is done (at least 5 different levels of humidity in the good range), click on the “Abort” button to stop the run before the completion of 100 cycles.

12. The results of the Humidity Calibration, called Humidity Calibration check, can be visualized using the Experiment Analysis tool.

Launching an Experiment

Different types of acquisition protocols can be launched according to the device and accessories associated.

As explained in the NeOse Advance section, make sure to warm up the device for a minimum of 30 minutes and to add non-sterile PTFE Luer filters on the Baseline inlets of both the NeOse Advance and the HeptaValve Mini. Also, make sure to warm up the Amplifier for a minimum of 30 minutes and to add non-sterile PTFE Luer filters on the Baseline inlet. 

Questioning Acquisition Protocol without sampler

The Questioning Acquisition Protocol allows for acquisitions of a single sample without using a HeptaValve Mini, nor attributing a name to the sample. It can be used for on-the-fly sample detection and recognition based on comparison with a previously built odor database. 

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
   
2. Select the device.
   
   
   
3. Select the protocol “Questioning acquisition (interactive)” and click on the "Next" button.
   
    
4. The run can be customized by modifying the name of the plan, the duration of analyte acquisition, the pump power and the time between acquisitions. The Questioning protocol includes only 1 cycle.
   
   
   
5. Tags can be added to apply to all samples by clicking on the "New tag" button, filling out the field and validated by hitting the return key. Added tags can be removed by clicking on the cross at right of the tag name.
   
   
   
6. Click on the “Run” button to start the experiment.
   
   
   
7. The Questioning acquisition protocol will start. Each step can be followed in the new window that opens: step of the measurement, number of cycles.
   Data can be viewed in real time by clicking on the “Live data view” button on the top right of the screen.
   
   
   
8. A pop-up window opens before the measurement starts to ensure that the user will present the sample to be measured. The sample vial must then be connected to the analyte inlet of the NeOse Advance. When ready, click on the “Yes” button.
   
   
   
   
   Delete

   Warning

    The protocol will not proceed to the next measurement unless the user clicks on the "Yes" button.

   
   
9. Once you click on the "Yes" button, a pop-up window will confirm the action and the protocol will continue.
   
   
   
10. Measurement of the baseline starts, followed by measurement of the sample. Once the sample is recorded, another pop-up window opens. Gently remove the sample from the NeOse Advance inlet and click on the "Yes" button to complete the experiment.
    
    
    
11. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen.
    
12. Click on the “Close” button to return to the Selection screen.
    
    
13. Disconnect the vial from the inlets by gently pulling the tubing out of each vial and removing the vent system.
    
    
14. Start a clean-up process as described in the section Cleaning protocol.
    
    
15. Data can be analyzed with the Experiment Analysis tool.

Generic Acquisition (unique sample) without sampler

The Generic Acquisition Protocol allows for automated data acquisitions of a single named sample without using an HeptaValve Mini. It can be used for on-the-fly sample detection and recognition based on comparison with a previously built odor database. 

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
2. Select the device and the protocol “Generic acquisition (unique sample)” and click on the "Next" button.
   
3. The run can be customized by modifying the name of the plan, the number of cycle, the duration of the analyte acquisition, the pump power and the time between acquisitions. This protocol includes only 1 sample.
   
   
   
4. The sample must be named. Tags can be added by clicking on the "New tag" button, filling out the field and validated by hitting the return key. 
   
   
   
5. Click on the “Run” button to start the experiment.
   
   
6. The Generic Acquisition Protocol (unique sample) will start. Each step can be followed in the new window that opens: step of the measurement, number of cycles. Data can be viewed in real time by clicking on the “Live data view” button on the top right of the screen.
   

   
   

   You can also open the Live Data View and the Annotator View windows by clicking on the “Live Sensor View” button and on the “Annotator View” button from the dedicated window.  The Live Data View and the Annotator View windows will be opened in your internet browser. 
   

   
   

   7. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen. 
   
   

   
   

   8. Click on the button “Close” to return to the home page.

   
   

Generic Acquisition (interactive) without sampler

The Generic Acquisition protocol interactive allows measurement of up to 7 different samples in a run without using an HeptaValve Mini. Data can be analyzed in the Visualization, Recognition or Detection reports.

The absence of an external sampler requires the operator to present and remove each sample at the right time. This interactive protocol guides the operator step by step.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
2. Select the device and the protocol “Generic acquisition (interactive)” and click on the "Next" button.
   
   
   
3. The run can be customized by modifying its name of the plan, the number of cycle, the duration of the analyte acquisition, the pump power, the purge sample line time and the time between acquisitions. 
   
   
   
4. Samples can be added by clicking on the “+ Add sample” button. A maximum of 7 samples can be recorded. You must label each sample, and can add information with individual tags.
   
   
   
5. Prepare the samples to be presented in the same order than indicated in this page from the top to the bottom.
   
   
6. Click on the “Run” button to start the experiment.
   
   
7. The protocol will start. Each step can be followed in the new window that opens. The name of the sample being measured is highlighted with the blue rolling circle and the total number of cycles with the current cycle number appearing on the right.
   
   
8. It is possible to monitor the evolution of several parameters in real time by clicking on the “Live Data View” button.
   
   You can also open the Live Data View and the Annotator View windows by clicking on the “Live Data View” button and on the “Annotator View” button from the dedicated window. The Live Data View and the Annotator View windows will be opened in your internet browser. 
   
   
9. Place your sample under the Sample inlet of the device. A message appears before each measurement to ensure that the user has changed the sample to be measured by placing a different vial under the inlet of the NeOse Advance. When the sample is ready, click on the “Yes” button.
   
10. When the sample intake is completed, a message appears at the end of each measurement. Remove your sample from the Sample inlet to ensure that there is no additional sample taken in. When the sample is removed, click on the “Yes” button.
    
11. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen. 
    
12. Click on the button “Close” to return to the home page.
    
    

Generic Acquisition Protocol with sampler

The Generic Acquisition protocol allows measurement of up to 7 different samples in a run. It requires a HeptaValve Mini connected to the NeOse Advance. Data can be analyzed in the Visualization, Recognition or Detection reports.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
   
2. Select the device.
   
   
   
3. Select the protocol to run “Generic acquisition (with Sampler)”. Click on the "Next" button.
   
   
4. Run parameters can be customized in the following window, e.g. number of cycles,  acquisition duration and labels for each sample.
   
   
   The name of the experiment can be changed to suit your need of traceability and simpler data browsing.The sample specifics are displayed on the right sequentially from Sample 1 (connected to HeptaValve Mini inlet 1) through Sample 7 (connected to HeptaValve Mini inlet 7).
   
   
   
5. Tags can be added to all samples by clicking on the "New tag" button, filling out the field and validated by hitting the return key. Added tags can be removed by clicking on the cross at right of the tag name.
   
   
   
6. Independent tags can be added for each samples.
   
   
   
7. To measure more samples, click on the “Add sample” button. A maximum of 7 samples can be recorded.
   
   
8. Prepare and connect the samples to the HeptaValve Mini, ensuring that the inlet number matches the samples defined in the run parameters. Tubing from the Heptavalve Mini is placed inside of the vial through the septum and a vent system is added. For more precise instruction, refer to the HeptaValve section of the User Manual.
   
   
9. Click on the “Run” button to start the experiment.
   
   
10. The Generic acquisition protocol will start. Each step can be followed in the new window that opens: sample being measured, step of the measurement, number of cycle.
    
    
    
11. Data can be viewed in real time by clicking on the “Live data view” button on the top right of the screen.
    
    
12. Once the experiment is completed, the confirmation of the end of the experiment appears.
    
    
    
13. Click on the “Close” button to return to the home page.
    
    
14. When the protocol is completed, disconnect the vials from the inlets by gently pulling the tubing out of each vial and removing the vent system. Disconnect the NeOse Advance from the HeptaValve Mini.
    
    
15. Start a clean-up process as described in the section Cleaning protocol section.
    
    
16. Results can be visualized with the Experiment Analysis tool.
    
    
    
    Delete

    Info

    To Abort before the end of the experiment, read the Stopping the Experiment paragraph instructions.

Limit of Detection Protocol with sampler

The Detection protocol is dedicated to defining the limit of detection (LOD) of an analyte in a diluent displaying a low or high background. It requires a HeptaValve Mini connected to the NeOse Advance. In addition to varied dilution levels, a sample of the pure analyte as well as the pure diluent are necessary for this protocol.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
   
2. Select the device.
   
   
3. Select the protocol  “Detection acquisition (with Sampler)” and click on the "Next" button.
   
   
4. Parameters of the run can customized. Default settings include 50 cycles, the diluent name defined as “Citroflex,” the analyte name defined as “Linalyl-Acetate”, and generic acquisition duration, pump power and time between acquisition for a common dilution series.
   
   The name of the experiment can be changed to suit your need of traceability and simple data browsing.
   
   
   
5. Tags can be added to all samples by clicking on the "New tag" button, filling out the field and validated by hitting the return key. Added tags can be removed by clicking on the cross at right of the tag name. 
   
   
6. Pure diluent sample must always be positioned on inlet 1 of the Heptavalve Mini. The different dilution levels and the pure analyte (corresponding to 100%) can be positioned anywhere in the HeptaValve Mini setup from inlet 2 to inlet 7 (respectively named Sample 1 and Sample 6 in the Acquisition Module).
   To measure more than 4 different dilution levels, click on the “Add sample” button. A maximum of 5 intermediate dilution levels can be recorded as one of the 7 available sample inlets is used for pure diluent (inlet 1) and one for pure analyte.
   
   
7. Prepare and connect the samples to the HeptaValve Mini, ensuring that the inlet number matches the dilution level defined in the run parameters. Tubing from the HeptaValve Mini is placed inside of the vial through the septum and a vent system is added. For more precise instruction, refer to the HeptaValve section of the User Manual.
   
   
8. Click on the “Run” button to start the experiment.
   
   
9. The Detection acquisition protocol will start. Each step can be followed in the new window that opens: sample being measured, step of the measurement, number of cycle.
   
   
   
10. Data can be viewed in real time by clicking on the “Live data view” button on the top right of the screen.
    
    
11. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen.
    
    
    
12. Click on the “Close” button to return to the home page.
    
    
13. When the protocol is completed, disconnect the vials from the inlets by gently pulling the tubing out of each vial and removing the vent system. Disconnect the NeOse Advance from the HeptaValve Mini.
    
    
14. Start a clean-up process as described in the section Cleaning protocol section.
    
    
15. Results can be visualized with the Experiment Analysis tool. .

Amplified generic acquisition (interactive) without sampler

The Amplified generic acquisition protocol without sampler allows measurement of up to 7 different samples in a run. It requires an Amplifier connected to the NeOse Advance. The records can be processed in the Amplifier post-analysis modes.

The absence of an external sampler requires the operator to present and remove each sample at the right time. This interactive protocol guides the operator step by step.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button. 
   
   
2. Select the device and the protocol to run “Generic acquisition (interactive with Amplifier)”. Click on the "Next" button.
   
   
   
3. All fields with a red star are mandatory. The run name and number of cycles can be customized.
   
   
   
4. All parameters regarding the behavior of the Amplifier (acquisition time, desorption temperature...) can be customized within the limits defined in the information pop-up.
   
   
   
5. Samples can be added by clicking on the “+ Add sample” button. A maximum of 7 samples can be recorded. Each sample must be named and additional information can be provided with individual tags.
   
   To prevent any confusion during acquisition and subsequent data analysis, samples should be presented by the user in the sequence listed and displayed on the right part of the screen.
   
   
6. Tags can be added to be applied to all samples.
   
   
   
7. Prepare the samples to be presented in the same order than indicated in this page.
   
   
8. Click on the “Run” button to start the experiment.
   
   
   
9. The protocol will start. Each step can be followed in the new window that opens. The name of the sample being measured is highlighted with the blue rolling circle and the total number of cycles with the current cycle number appearing on the right.
   
   
10. It is possible to monitor the evolution of several parameters in real time by clicking on the “Live Data View” button.
    
    
    You can also open the Live Data View and the Annotator View windows by clicking on the “Live Sensor View” button and on the “Annotator View” button from the dedicated window. The Live Data View and the Annotator View windows will be opened in your internet browser. 
    
    
11. Place your sample under the Sample inlet of the Amplifier. A message appears before each measurement to ensure that the user has changed the sample to be measured and placed a different vial under the Sample inlet of the Amplifier. When the sample is ready, click on the “Yes” button.
    
    
    
12. When the sample intake is completed, a message appears at the end of each concentration step. Remove your sample from the Sample inlet to ensure that there is no sample taken in during purge. When the sample is removed, click on the “Yes” button.
    
    
    
13. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen. 
    
    
    
14. Click on the button “Close” to return to the home page.
    
    

Amplified acquisition in open environment without sampler

This protocol is fully automatic and allows the user to concentrate and measure a sample periodically for a preset number of times, with no interaction with the user. It is mostly dedicated to low concentration samples (under the limit of detection) or to odors from large, open environments. Purge will thus not be required in this mode.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
2. Select the device and the protocol to run “Open environment acquisition (with Amplifier)”. Click on the "Next" button.
   
3. In the following window, the run can be customized. The number of cycles and the acquisition duration fields must be filled out.
   All Amplifier’s parameters can be customized. Tags can be added.
   
   
   
4. Sample must be named. If necessary, additional individual tags can provide more information. Click on the “+ Add sample” button to add replicates.
   
   
   
5. Click on the "Run" button to start the experiment.
   
   
6. The protocol will start. Each step can be followed in the new window that opens. The name of the sample being measured is displayed next to a waiting icon and the total number of cycles with the current cycle number appearing on the right.
   

   
   

7. It is possible to monitor the evolution of several parameters in real time by clicking on the “Live Data View” button.
   
   You can also open the Live Data View and the Annotator View windows by clicking on the “Live Sensor View” button and on the “Annotator View” button from the dedicated window.  The Live Data View and the Annotator View windows will be opened in your internet browser. 
    

8. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen. 
   
   

9. Click on the button “Close” to return to the home page.

   
   

Amplified generic acquisition with sampler

The Amplified generic acquisition protocol allows measurement of up to 7 different samples in a run. It requires both an Amplifier and a HeptaValve Mini connected to the NeOse Advance. The records can be processed in the Amplifier post-analysis modes.

This protocol is fully automated, no human interaction is required once the run is started.

1. Open the Acquisition module by clicking on the "Start a new acquisition" button.
   
   
2. Select the device and the protocol to run “Generic acquisition (with Amplifier & Sampler)”. Click on the "Next" button.
   
   
   
3. All fields with a red star are mandatory. You can customize the run name, and the number of cycles.
   
   
   
4. All parameters regarding the behavior of the Amplifier, (acquisition time, desorption temperature...) can be customized within the limits defined in the information pop-up.
   
   
   
5. You can add samples by clicking on the “+ Add sample” button. A maximum of 7 samples can be recorded. Each sample must be named, and additional information can be provided with individual tags.
   

   

   
   The sample labels on the right are displayed in order from inlet 1 through 7. 
   
   
6. Tags can be added to be applied to all samples.
   
   
   
7. Prepare and connect the samples to the HeptaValve Mini, ensuring that the inlet number matches the samples defined in the run parameters. Tubing from the Heptavalve Mini is placed inside of the vial through the septum and a vent system is added.
   
   
8. Click on the "Run" button to start the experiment.
   
   
9. The Generic acquisition protocol will start. Each step can be followed in the new window that opens. The name of the sample being measured is highlighted in dark grey and the total number of cycles with the current cycle number appearing on the right.
   
   
   
10. It is possible to monitor the evolution of several parameters in real time by clicking on the “Live Data View” button.
    
    You can also open the Live Data View and the Annotator View windows by clicking on the “Live Sensor View” button and on the “Annotator View” button from the dedicated window. The Live Data View and the Annotator View windows will be opened in your internet browser. 
    
    
    

11. Once the experiment is completed, the confirmation of the end of the experiment appears on the screen. 
    
    

12. Click on the button “Close” to return to the home page.

13. Disconnect the vial from the inlets by gently pulling the tubing out of each vial and removing the vent system and disconnecting the Amplifier from the HeptaValve Mini.

14. Start a clean-up process.

End of experiment

After each measurement session a thorough system cleaning, as explained in section Cleaning protocol, must be performed to avoid a contaminated system for the next experiment. Aryballe does not guarantee results from experiments performed on dirty NeOse Advance system.

For system cleaning, the NeOse Advance needs to be disconnected. Only the HeptaValve Mini and the external pump box are necessary. During this procedure, make sure that no samples are connected to inlets.

System Shutdown

When the device is clean, shut down the system:

1. Close the Acquisition module window.
   
   
   
2. Close (if opened) the Live Data View and Annotator windows in the browser.
   
   
3. Unplug the NeOse Advance from the computer by removing the USB cable.
   
   
4. Put the transparent caps back on both inlets of the NeOse Advance.

Stopping an Experiment at any time 

1. Click on the “Abort” button to stop the run before the completion of the whole experiment.

2. A pop-up window will open. Click on the “Abort” button to validate that you want to abort the experiment.

3. The confirmation that you have requested to abort the experiment will appear in the main window.

4. The experiment will abort once the measurement currently in process is completed. Click on the "Close" button to return to the Acquisition screen.

Experiment Analysis Tool

The Experiment Analysis tool creates reports for the visualization and analysis of experiments. The humidity correction and/or the drift correction can be applied on the experimental run via this tool.

Before making any statistical analysis, the performance of the device, setup and the calibrations must be verified.

Quality Check

Report Selection

1. In the Aryballe Suite home page, click on the "Consult analysis and reports" button.
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
2. The Experiment Analysis window will open. Click on the “Quality check” button.
   
   
   
3. Select the Quality check run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
   
   
4. Click on the “Check” button to generate a report.
   
   
   
5. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Quality.

Report Structure

It is possible to select several Quality check runs in the same analysis. The results will be presented run by run in different tabs.

For a selected run, a tab contains:

1. The intensity graph for each cycle and each compound (average and maximum values).
   

    

2. The CQS score analysis with:
   • The Global CQS, i.e. the score representing the quality of global separation of samples. On a set of N different clustered odors, it is represented only by one CQS score.
   • The Individual CQS, the score representing the quality of each individual cluster (which contains multiple measurements coming from one same sample) among the other clusters. On a set of N different clustered odors, it is represented by N different numbers.
     
     
     A CQS score superior to 50% confirms optimal performance of the sensor.
     
     
     
3. A PCA analysis:
   
   Principal Component Analysis (PCA) is a statistical tool used for multivariate data analysis. In our case, the normalized response of the array of n-biosensors is projected and reduced in a new n-space where the first components explain the majority of the variance between records. Each marker is an acquisition, and each color is a sample. The data points on the two first dimensions plot that are located close to each other indicate a degree of odor proximity, while distant dots indicate different sample odors.
   
   A good performance of the device is indicated by markers of a same color tightly clustering together, while markers of different colors should be clearly separated in different clusters.

System Cleanliness Check of the sampler

Report Selection

1. In the Aryballe Suite main page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Cleanliness check” button.
   Choose the Cleanliness check run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
   
   
   
3. Click on the “Check” button to generate a report.
   
   
   
4. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Cleanliness.
   
   

Report Structure

The Cleanliness check report is made of as many tabs as there are cycles. One tab represents the measurements of one cycle for all the inlets of the HeptaValve Mini.

The row of tabs allows you to select the cycle. For each cycle, a colored graph represents an inlet.

Each graph is divided in two parts:

• Left is the baseline, delimited by two vertical grey lines.
• Right is the injection of the selected sample, delimited by two vertical blue lines.

The chart represents the average intensity value across all the biosensors and the colored areas delimit the maximum and minimum values.

Validation of the setup’s cleanliness:

• If the odor intensity signal presents a peak upon valve switching (see example below), then the system is dirty and another cleaning session needs to be done
  
• Once the system is clean, the odor intensity signal should be stable (presented as minor fluctuations with very small intensity value, typically < 0.2 as shown on the Y axis of the chart).
  

  
  

Humidity Calibration Check

Report Selection

1. In the Aryballe Suite home page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Humidity calibration check” button to visualize the calibration.
   
   
3. Choose the Humidity calibration check run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
   
   
4. Click on the “Check” button to generate a report.
   
   
   
5. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Humidity.
   
   

Report Structure

The report contains a tab for each selected calibration run, which displays a single Humidity report figure. For each spot, this report displays the relation between relative humidity and signal intensity for each biosensor.
The report allows to verify that the humidity calibration went well and that there is no anomaly in the behavior of biosensors.

Visualization Analysis

Report Selection

1. In the Aryballe Suite home page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Visualization” button to visualize the list of all the runs.
   
   
3. Choose the run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
   
   
4. Humidity correction can be added to the analysis. Check the boxes at the left of the Humidity Calibration run of interest in the dedicated tab.
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
5. Click on the “Visualize” button to generate a report.
   
   
6. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Visualization. The .csv documents corresponding to the report are saved automatically at the same location.
   
   
   

Report Structure

At the top of the report, the title, the device name and run id are displayed.

The Visualization report contains the following tabs (described in the sections below):

• Signature
• Evolutions
• Data, displaying PCA charts and Mean relative error matrix
• Sensors
• Time drift
• Humidity correction (if one is applied)
• Drift correction (if one is applied)
  
  

Signature: several types of odor signatures are displayed as a radar chart.

• Measured: All raw signatures for every sample and every cycle
• Measured (simplified): Average raw signature over the cycles for each sample (with +/- standard deviation per spot envelope)
• Normalized: All normalized signatures for every sample and every cycle
• Normalized (simplified): Average normalized signature over the cycles for each sample (with +/- standard deviation per spot envelope)
  
  

Evolutions: Evolution of signature average (with min and max of each signature envelop) over time for each sample. This will include humidity correction if it is applied.

Data (PCA + Distance Matrix): Principal Component Analysis (PCA) is a statistical tool used for multivariate data analysis. In our case the normalized response of the array of n-biosensors is projected and reduced in a new n-space where the firsts components explain the majority of the variance between records.

Each marker is an acquisition, and each color is a sample. The data points on the two first dimensions plot that are located close to each other indicate a degree of odor proximity, while distant dots indicate different sample odors. 

PCAs allow to quickly visualize separation of samples. Distant records (represented by a dot) indicate that records are different, while records indistinguishable or close from each other indicate that the measures have an odor proximity. The global CQS, shown on top of the PCA chart indicates how well all samples are separated from each other, while individual CQS shown in the legend at the right of the graph represents how the sample is differentiated form the other samples in the PCA.

The distance matrix is based on the calculation of the differences between the reaction of all biosensors that were activated during measurement and give the distance value between two records. These numeric values indicate the degree of difference between samples.

Sensors: Displays humidity (top) and temperature (bottom) evolutions over the experiment duration, i.e. the cycles for the baseline and analyte portion of the records. This allows a simple view of the delta in humidity and temperature between the baseline and analyte over time. 

Time Drift: The first graph displays the evolution of the average baseline value (averaged on all spots with +/- standard deviation envelope) per record, the first baseline value being set to zero. The second graph displays, per compound, the mean relative error between the current signature and the first signature of the dataset.

Humidity Correction: Results with corrections applied are displayed as PCA, distance matrix and CQS. are presented in a dedicated tab. This allows the user to toggle between the corrected/uncorrected data to see the impact of these factors on the results.

Detection Analysis

Report Selection

1. In the Aryballe Suite home page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Detection” button to visualize the list of all the Detection runs.
   
   
3. Choose the learning run(s) which will be used to build the model by checking the boxes at the left of the run dates. Aryballe’s processing and learning pipeline will learn to detect the analyte in the diluent. The number of runs selected will appear in the lower bar.
   
   
   
4. A test run can be chosen to validate the quality and the robustness of the model.
   
   
   
5. Select a questioning run.
   
   
   
6. Humidity correction can be added to the analysis. Check the boxes at the left of the Humidity Calibration run of interest in each dedicated tab.
   
   
7. Click on the “Analyze” button to generate a report.
   
   
   
8. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Detection. The .csv documents corresponding are saved automatically at the same location.
   
   

Report Structure

The first tab is a summary of the report (see in purple below). Then the other ones, correspond to the Learning, the Testing and the Questioning, respectively called L, T or Q. The “*” is used to present the results with Humidity correction.

Each tab displays the following figures:

• Run info (number of cycles, analytes, etc)
• LOD results

The goal of the LOD algorithm is to build a similarity index in order to find a threshold to separate the pure diluent from the other dilution rates. The similarity index is computed using the signature of the pure compound and the signatures intensities. The threshold is set so that 5% of the pure diluent population is above it (5% of false positive rate). Then, the detection rate, per dilution rate, is computed as the rate of the population that is above this threshold.

A table gives a resume of the detection rate per dilution rate.

The confusion matrix is a metric to determine the quality of how the NeOse Advance (augmented with machine learning) recognizes, or not, the presence of the anlayte molecule. This metric is computed by presenting a sufficiently high number of “test” odors (for instance 100 for each class), then providing prediction of their nature based on their signatures and thanks to a pre-learned model. All the responses are used to fill the confusion matrix arranged in column/line as actual/predicted responses. Once normalized with probabilities % of responses (instead of counts), the identity matrix represents the perfect recognition. Here, only two categories are distinct: the “pure diluent” or the “molecule + diluent”.

The Confusion Matrix is presented for the Learning and the Testing phase but not for the Questioning because the records are not named.

Recognition Analysis

Report Selection

1. In the Aryballe Suite home page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Recognition” button to visualize the list of all the runs.
   
   
3. Choose the learning run to analyze by checking the box at the left of the run dates. Only one run can be selected for the learning phase. The number of runs selected will appear in the lower bar. 
   
   
   
4. One or several testing runs can be chosen in order to validate the quality and the robustness of the model.
   
   
   
5. Select one or several questioning runs.
   
   
   
6. Click on the “Analyze” button to generate a report.
   
   
   
7. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/ Recognition. The .csv documents corresponding are saved automatically at the same location.
   
   

Report Structure

The organization of the recognition report is similar to that of the detection report.

The first tab is a summary.  Other ones correspond to the Learning, the Testing and the Questioning, respectively called L, T or Q. The “*” is used to present the results with a Humidity correction. 

Each tab displays the following figures:

• PCA
  Principal Component Analysis (PCA) is a statistical tool used for multivariate data analysis. In our case the normalized response of the array of n-biosensors is projected and reduced in a new n-space where the firsts components explain the majority of the variance between records.
  Each marker is an acquisition, and each color is a sample. The data points on the 2 first dimensions plot that are located close to each other indicate a degree of odor proximity, while distant dots indicate different sample odors.
• Confusion matrix
  The confusion matrix is a metric to determine the quality of how the sensor (augmented with machine learning) recognizes a specific set of odors. This metric is computed by presenting a sufficiently high number of “test” odors (for instance 100 for each class), then providing prediction of their nature based on their signatures and thanks to a pre-learned model. All the responses are used to fill the confusion matrix arranged in column/line as actual/predicted responses. Once normalized with probabilities % of responses (instead of counts), the identity matrix represents the perfect recognition.

Amplifier cleanliness check

Report selection 

1. In the Aryballe Suite main page, click on the "Consult analysis and reports" button.
   
   
   
2. The Experiment Analysis window will open. Click on the “Amplifier cleanliness check” button.
   
   Choose the Cleanliness check run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
3. Click on the “Generate” button to generate a report.
   
   
4. The report will automatically open in a new browser window and will be saved in the computer under Documents/Aryballe/Amplifier.
   
   

Report Structure

The Cleanliness check report is made of one Summary and one Amplifier Clean Check tab where you can visualize one graph for each cycle. 

Each graph is divided in two parts:

• Left is the baseline, delimited by the dark discontinous line.
• Right is the thermodesorption measure

Validation of the setup’s cleanliness:

• Without concentration, the thermodesorption signal should be unsignificant as below : 

•  If the peak of thermodesorption is signiticant, meaning that the Tenax is polluted or could had stock VOCs by diffusion during the storage of the Amplifier, redo a cleanliness check of at least 2 cycles.

Amplifier report analysis

Report selection 

1. In the “Aryballe Suite for Amplifier Public” directory, double click on the Experiment Analysis icon.
   
   
   
2. The Runner Experiment Analysis window will open. Click on the “Amplifier” button to visualize the list of all runs.
   
   
   
3. Choose the run(s) you want to visualize by checking the boxes at the left of the run dates. The number of runs selected will appear in the lower bar.
   
   
   
4. Click on the “Generate” button to generate a report.
   
   
5. The report will automatically open in a new browser window shortly after it is generated and will be automatically saved in the computer under Documents/Aryballe/Amplifier.
   
   

Report structure

For selected runs, the Amplifier report provides a Summary tab and as many tabs as there are individual samples. HTML reports are displayed on a web interface such as Mozilla Firefox or Chrome.

Reports are semi-dynamic: you can navigate between panels that will inform you on the discrimination of the records you made, or the intensity of those measurements.

The report window presents the following panels:

Summary : This main panel summarizes the main experimental results.

On the top left chart, the intensities of all the records are displayed, grouped by nature of sample. The intensity appears in radians (rad) and corresponds to the average response across biosensors at the apex of the thermo desorption curve.

The top right chart has a similar organization and presents the area under the selected zone of peak detection. The bottom left chart is a Principal Component Analysis (PCA) representation of the run. Each nature of sample gets its own color, and each sample analyzed gets its own dot on this 2-D chart. The x axis presents the projection of the olfactive signature on the first principal component of the PCA (PC1), and the percentage of expressed variance (59,6% in this example). Similarly, the y axis presents the projection of each record on the second principal component of the PCA (PC2). On this chart, points that appears close have measured olfactive signature that are perceived similar by the sensor, when the furthest points generate the most different signatures.

To ease the reading of this discrimination tool, the bottom right chart presents the global Clustering Quality Score: it is a numerical approach to quantify the performance of discrimination. This value is based on inter- /intra-cluster distances and provides an information on how well samples are clustered and thus discriminated.

You can then further navigate in the individual sample report to see the details of each record by clicking on the name of your sample, where you can then see each thermal desorption cycle per cycle. 

The first temporal data is called a sensorgram: it gives information of the variation of intensity of each biosensor during time. The highlighted green zone is determined automatically by an algorithm and frames the extraction of an olfactive signature, presented in the bottom right. The vertical dotted line indicates the switch of the valve in the Amplifier, corresponding to the beginning of the flow of the adsorbed sample towards the sensor. The second temporal chart presents a red curve corresponding to the temperature of the thermo-desorber, and a blue one recording the amount of humidity during the thermal desorption. A humidity-specific sensor is placed online just after the odor sensor and will monitor the variations of relative humidity of the sample trapped during the concentration. The tables appearing on the top right of the page are numerical informations about the record:

• Apex gives the average intensity of all the biosensors reached on the peak.
• FWHM (Full Width at Height Maximum) gives an information on the width of the peak and is given in seconds.
• Tr is the retention time corresponding to the time to reach the maximum of signal of the peak (seconds). Value is taken after the switch of the injection valve (dotted lines on the chart) and can be characteristic about a chromatographic effect of the peak.
• Area is extracted under the blue zone and is another way to quantify the amount of signal than the Apex value.

Miscellaneous

Log-out procedure

1. Click on Windows Start Button
2. Use the built-in search bar to find the Credentials Manager, then open it
3. Choose Windows Credentials
4.  
5. In the list of credentials, look for “Aryballe-Suite/<username>” field. In particular case, <username> will be replaced by your account's username, e.g. "Aryballe-Suite/John"
6. Use the Remove button to delete the credential.
7. Close the Credential Manager 
8. Restart the Aryballe Suite application and log in again
9. You can now use the Aryballe Suite application using the newly logged in account

Offline Authentication

The Aryballe Suite may be used in an Offline Mode to conform with your organization's IT requirements. Use these instructions if online authentication is not possible on your PC or in your organization. The procedure is to be performed every time you log out of the application.

Prerequisites

• Token file provided by Aryballe must stored on the computer's local drive.
• Aryballe Suite version v2021.12.09 or higher must be installed.

Procedure

1. Open Windows Start Menu
2. Open the Aryballe Suite (advanced) folder 
3. Select Maintenance from the list
   
4. Drag and drop the Token file onto the Register offline license item (script)
5. A console window will pop up with the following message. If there is an error message, please contact support@aryballe.com for assistance.
6. Close the console window. You are authenticated with an offline token and can use the Aryballe Suite in offline mode.