Franklin implements technical and procedural safeguards to uphold the integrity, reliability, and completeness of data processing. Automated alerts are configured to identify and surface deviations, incomplete validations, and critical system changes that may impact the accuracy or validity of the analysis. These controls are embedded within operational and development workflows to ensure early detection of potential issues and minimize manual oversight.
This article compiles all alerts, notifications, and built-in technical controls and safeguards that users should be aware of, providing clear visibility into the system’s checkpoints.
Franklin includes a series of automated safeguards that operate at key points in the user workflow. These safeguards are designed to prevent data entry errors, enforce input consistency, alert users to potential issues, and ensure that critical actions are confirmed before execution.
Login Activity Monitoring
Login activity is monitored and alerts are generated for events such as failed login attempts, unexpected access times, or access from unrecognized locations. These alerts help ensure that only authorized personnel interact with the system.
Note: For detailed information about authentication methods, password policies, and role-based access, refer to the Access Control and Authentication article.
Search Field Input Validation
Alerts are triggered by the user’s input in case of an unsupported or incorrect format (e.g., missing genomic coordinates), preventing downstream errors by enforcing consistent input structure at the point of entry. Examples of validation alerts include:
Unsupported format: If the variant notation cannot be parsed by the system, a message is displayed: “Franklin couldn’t parse the provided variant, please check the format examples.”
Wrong reference: If the reference allele at the specified position does not match the provided input, the system displays: “Wrong ref. Ref at this location is [correct allele].”
Missing gene: If a transcript identifier is submitted without an associated gene, the alert “Search is missing a gene” is displayed.
These validations ensure that variant searches conform to the expected nomenclature and reference data before any downstream analysis is initiated.
User Validation of Transcript
When multiple transcript options are available for a given variant, users are required to actively select the clinically relevant transcript. A dialog box is presented with the message “Please Select the Variant You’re Referring to” and lists the available transcript-variant combinations (e.g., chr1:94564383 A>T, NM_000350). This safeguard ensures that variant classification and downstream interpretation are based on the most appropriate transcript for the clinical context.
Important: Selecting an incorrect transcript may lead to misclassification. Users should verify transcript selection against their organization’s clinical protocols.
Sample Sheet Upload and Validation
Users can upload a sample sheet in XLS, XLSX, or CSV format. If the upload fails (e.g., due to an internet outage), a notification will appear in the UI. In such cases, users should verify their internet connection and confirm that the file format is valid (XLS, XLSX, or CSV). If the issue persists, users should contact Franklin Support for further assistance.
Once the file is uploaded successfully, validation is performed on the sample sheet metadata—not the associated files. The validation outcomes are as follows:
No warnings or errors: The sample sheet is accepted, all cases are created, and a success message is shown in the UI.
Warnings only: The upload is completed and all cases are created. An email is sent to the uploading user detailing each warning, including the line, column, and data item involved (e.g., line 55 – ethnicity ‘test 123’ doesn’t exist). Users should review and correct the sample sheet accordingly.
Errors detected: The upload completes but no cases are created. All errors and warnings are shown in the UI. A detailed email is sent to the uploading user with the line, column, and data item causing the failure (e.g., line 55 – assay ‘TSO 501’ doesn’t exist). Users must update the sample sheet based on the error reasons.
Post-Validation Issues:
Once validation is completed, the system proceeds to create the samples and their linked cases. If an issue occurs during this stage, a generic error message may appear without specific details. Users should re-upload the sample sheet. If the issue persists, they should contact Franklin Support.
Cases that are successfully created will appear in the system with the status "Pending Samples".
Sample-to-File Linking
The sample-to-file linking process acts as a safeguard by ensuring that only complete and correctly named data files are associated with each sample before any downstream analysis begins.
During the sample sheet upload, users must specify the S3 path or the basespace project and samples where the sample files are stored and the sample name for each entry.
For S3, any files in the provided folder that are prefixed with the sample name are automatically linked to the corresponding sample.
For basespace, any files found in the biosample or the appsession are linked to the sample.
If the exact set of required files is known when defining a workflow, the expected file composition should be configured in the assay definition. Linking will not proceed until all required files are detected in the specified S3 path.
Once the sample sheet is successfully uploaded, the system begins checking for file availability immediately. Subsequent checks are performed every 5 minutes, ensuring timely linking once all requirements are fulfilled.
Missing Sample Warning
This alert prevents users from proceeding with incomplete data and ensures corrective action is taken before analysis or reporting begins, supporting traceability and diagnostic integrity. Cases with missing or unlinked samples will display the status “Pending Samples” in the case list.
Confirmation Prompts
Franklin includes confirmation prompts (“Are you sure?” pop-ups) in scenarios where user actions may result in irreversible changes or data loss. These prompts are triggered before critical operations such as deleting records, submitting finalized analyses, or navigating away from unsaved content. The purpose is to ensure that users explicitly acknowledge the consequences of their actions, particularly when changes cannot be undone or when unsaved edits will be discarded. This mechanism helps prevent accidental data modification and reinforces informed decision-making.
Environment Mismatch Alert on Login
To prevent data access issues and ensure proper environment segregation, an alert is shown when users log into the community environment instead of their assigned organizational environment. The dialog prompts: “Are You in the Right Franklin Environment?” and provides a direct link to the correct environment (e.g., enterprise.genoox.com) with a “Redirect Me” button. This safeguard prevents users from inadvertently working in the wrong environment, which could affect data visibility, access controls, and audit trail integrity.
Session Timeout
For organizations using SSO, users are notified when their session expires and prompted to sign in again through their identity provider. The dialog displays: “You’ve Been Logged Out – It seems like you’ve been inactive for a while. To protect your data, we’ve ended your session.” Users can click “Log Me Back In” to re-authenticate. This ensures that only authenticated users maintain access and reduces the risk of unauthorized access from unattended sessions.
Caution: For accounts not using SSO, there is no default system timeout. Users should manually log out after completing their work or when stepping away from their workstation for an extended period.
Quality Control Monitoring
Franklin quality control (QC) metrics are displayed in the Workbench screen as a separated tab. These metrics include coverage metrics, sequencing quality, sex concordance, and relatedness checks. QC thresholds are assay-specific and configurable, allowing users to define hard or soft warnings based on predefined criteria. Each assay is monitored appropriately and potential quality issues are flagged before interpretation begins.
QC metrics include, among others: Average Variant Depth, Hom/Het Ratio, Ti/Tv Ratio, Variant Quality, Number of SNPs, Duplication Check, and Sex Detection. When QC warnings are detected, they are displayed prominently in the Quality Control tab for user review.
Mandatory Fields
Franklin enforces mandatory fields across key workflows to ensure completeness and standardization of critical data. These required fields must be filled before users can proceed. For example, when creating a new case, the Case Name field is mandatory and displays the validation message “Case Name is required” if left blank. This prevents submission of incomplete records and ensures that all necessary metadata is captured at the point of entry.
Case Processing Notifications
A notification is displayed in the Case Workbench when a case is still processing. This alert indicates that the analysis is not yet complete and additional time is needed, helping users avoid acting on incomplete or partial results. The message “Processing – We are evaluating confidence and prioritization for each variant” is shown until the analysis pipeline completes.
Important: Do not proceed with interpretation or reporting until processing is complete. Acting on partial results may lead to incorrect clinical conclusions.
Note: For information about encryption standards, infrastructure security, and network controls, refer to the Minimum Secure Operating Environment and Network and Connectivity Guidance articles.