JP-7857410-B2 - To facilitate the secure execution of external workflows for genome sequencing diagnostics.

JP7857410B2JP 7857410 B2JP7857410 B2JP 7857410B2JP-7857410-B2

Inventors

マイケル・スムート
ジェレミー・ウォード
アダム・フレンチ
ラッセル・テイラー

Assignees

イルミナインコーポレイテッド

Dates

Publication Date: 20260512
Application Date: 20220930
Priority Date: 20220926

Claims (20)

It is a system, A field-programmable gate array (FPGA) locally housed on a shared network server having a container orchestration engine and a variant analysis model, When executed by the FPGA, the system will Variant analysis is performed using the variant analysis model described above in order to generate nucleotide base calls for the sample nucleotide sequence, The diagnostic analysis is performed on the nucleotide base call to detect the genetic state of the sample nucleotide sequence by utilizing the FPGA to access the nucleotide base call in the shared network server and to sequentially execute processes for the diagnostic workflow container as scheduled by the container orchestration engine, wherein the diagnostic workflow container constitutes and executes an external sequencing diagnostic workflow from an external server without accessing the nucleotide base call in the shared network server. To prevent the exposure of the nucleotide base call to the external server, one or more workflow containers associated with each function of the external sequencing diagnostic workflow are isolated using targeted security authorizations. A system comprising: executing the external sequencing diagnostic workflow by utilizing the container orchestration engine to allocate computing resources of the FPGA to process one or more sequentially scheduled workflow containers; and a non-temporary computer-readable medium containing instructions to perform the above.
The system according to claim 1, further comprising instructions, when executed by the FPGA, causing the system to implement the external sequencing diagnostic workflow identified from an external application hosted on an external server separate from the shared network server of the variant analysis model and the container orchestration engine.
When executed by the FPGA, the system will Determining the diagnostic execution mode corresponding to the standardized genetic diagnostic protocol, The system according to claim 1 or 2, further comprising an instruction to a client device to grant it access only to diagnostic applications compatible with the diagnostic execution mode.
The system according to claim 1, further comprising instructions, when executed by the FPGA, causing the system to prevent the external sequencing diagnostic workflow from accessing the sequencing data associated with the nucleotide base call for the sample nucleotide sequence by applying read-only permissions to the workflow container used by the external sequencing diagnostic workflow.
The system according to claim 1, further comprising instructions, when executed by the FPGA , causing the system to selectively prevent one or more workflow containers from accessing sequencing data associated with the nucleotide base call for the sample nucleotide sequence while the external sequencing diagnostic workflow is being executed.
The system according to claim 1, further comprising, when executed by the FPGA, instructions causing the system to isolate the one or more workflow containers by utilizing the targeted security permissions in order to individually specify the sequence determination data access permissions for the one or more workflow containers.
The system according to claim 1 , further comprising, when executed by the FPGA, instructions causing the system to receive the external sequencing diagnostic workflow generated by an external device operated by an external entity.
A computer implementation method, To generate nucleotide base calls for the sample nucleotide sequence, variant analysis is performed using a variant analysis model, and The diagnostic analysis is performed on the nucleotide base call to detect the genetic state of the sample nucleotide sequence by utilizing an FPGA locally housed on the shared network server having the container orchestration engine and the variant analysis model, in order to access the nucleotide base call in the shared network server and to sequentially execute processes for the diagnostic workflow container as scheduled by the container orchestration engine, wherein the diagnostic workflow container constitutes and executes an external sequencing diagnostic workflow from an external server without accessing the nucleotide base call in the shared network server. To prevent the exposure of the nucleotide base call to the external server , one or more workflow containers associated with each function of the external sequencing diagnostic workflow are isolated using targeted security authorizations. A computer implementation method comprising executing the external sequencing diagnostic workflow by utilizing the container orchestration engine to allocate computing resources of the FPGA to process one or more sequentially scheduled workflow containers.
The computer implementation method according to claim 8, wherein separating one or more workflow containers includes controlling access to different workflow data sources for one or more workflow containers in order to prevent access to sequencing data by the external sequencing diagnostic workflow.
The computer implementation method according to claim 8 or 9, further comprising receiving a label index defining the version of the variant analysis model and a memory allocation used to perform the external sequencing diagnostic workflow.
Specifying multiple workflow data sources that store different types of workflow data, The computer implementation method according to claim 8, further comprising activating access to a first workflow data source among the plurality of workflow data sources for one workflow container from among the one or more workflow containers, while preventing access to other workflow data sources among the plurality of workflow data sources.
The computer implementation method according to claim 8 , further comprising mounting a plurality of workflow data sources as read-only for one or more workflow containers.
The computer implementation method according to claim 8, further comprising accessing the sequencing data associated with the nucleotide base call for the sample nucleotide sequence and preventing the external sequencing diagnostic workflow from accessing the sequencing data associated with the nucleotide base call for the sample nucleotide sequence by applying read-only permissions to the workflow container used by the external sequencing diagnostic workflow.
The computer implementation method according to claim 8 , further comprising performing the external sequencing diagnostic workflow during variant analysis using the variant analysis model.
A non-temporary computer-readable medium, when executed by an FPGA, to a computing device, To generate nucleotide base calls for the sample nucleotide sequence , variant analysis is performed using a variant analysis model, and The diagnostic analysis is performed on the nucleotide base call to detect the genetic state of the sample nucleotide sequence by utilizing the FPGA locally housed on the shared network server, which has the container orchestration engine and the variant analysis model, in order to access the nucleotide base call in the shared network server and to sequentially execute processes for the diagnostic workflow container as scheduled by the container orchestration engine, wherein the diagnostic workflow container constitutes and executes an external sequencing diagnostic workflow from an external server without accessing the nucleotide base call in the shared network server. To prevent the exposure of the nucleotide base call to the external server, one or more workflow containers associated with each function of the external sequencing diagnostic workflow are isolated using targeted security authorizations. A non-temporary computer-readable medium including instructions to perform the external sequencing diagnostic workflow by utilizing the container orchestration engine to allocate computing resources of the FPGA to process one or more sequentially scheduled workflow containers.
The non-temporary computer-readable medium according to claim 15, further comprising instructions, when executed by the FPGA, causing the computing device to isolate one or more workflow containers satisfying one or more standardized genetic diagnostic protocols while also executing the external sequencing diagnostic workflow, by encoding a workflow execution application that grants the external sequencing diagnostic workflow read-only access to the sequencing data associated with the nucleotide base call of the sample nucleotide sequence during the execution of the external sequencing diagnostic workflow.
The non-temporary computer-readable medium according to claim 15 or 16, further comprising instructions, when executed by the FPGA, causing the computing device to execute the external sequencing diagnostic workflow generated by an external system on a server separate from the shared network server of the container orchestration engine and the variant analysis model.
The non-temporary computer-readable medium according to claim 15, further comprising, when executed by the FPGA, instructions causing the computing device to isolate the one or more workflow containers by preventing one of the workflow containers from accessing the sequencing data of the sample nucleotide sequence.
The non-temporary computer-readable medium according to claim 18, further comprising instructions, when executed by the FPGA, causing the computing device to prevent the workflow container from accessing the sequence determination data by preventing the workflow container from accessing one or more workflow data sources, including an input directory, an output directory, and an application directory.
The non-temporary computer-readable medium according to claim 15, further comprising instructions that, when executed by the FPGA, cause the computing device to trigger the execution of the external sequencing diagnostic workflow by receiving post-defined parameters for implementing the external sequencing diagnostic workflow via the container orchestration engine.

Description

(Cross-reference of related applications) This application claims the benefit and priority of U.S. Patent Application No. 17/935476, filed on September 26, 2022, which claims priority to U.S. Provisional Patent Application No. 63/293587, filed on December 23, 2021. The aforementioned application is incorporated herein by reference in its entirety. In recent years, biotechnology companies and computer science institutions have improved hardware and software for generating diagnostics on nucleotide sequences of genomic samples. In particular, some existing diagnostic platforms generate nucleotide base calls from nucleotide reads of sample nucleotide sequences and/or perform diagnostics on nucleotide base calls for various purposes. For example, existing diagnostic systems perform diagnostic applications (e.g., cancer screening assays) to screen nucleotide sequences for cancer by detecting specific genetic markers within the nucleotide base calls of sample sequences. Some existing diagnostic systems also perform other diagnostics, such as genetic testing, to determine other genetic conditions (or predispositions to developing genetic conditions) or other genetic traits. Despite these recent advancements, existing diagnostic systems continue to exhibit numerous shortcomings or disadvantages. For example, many conventional diagnostic systems are rigidly tied to a specific set of internal diagnostic applications, limiting their scope and usefulness. Indeed, many conventional systems can only perform genomic diagnostics using applications specifically designed for and installed on the system. Therefore, when a bioinformatics practitioner requires a specific diagnostic analysis of nucleotide sequences, existing systems may be unable to provide the necessary analytical data if a platform-specific diagnostic application for the analysis has not yet been written and/or installed within the system. Apart from their lack of flexibility, some conventional diagnostic systems exhibit coding or network vulnerabilities that compromise or expose private or confidential genetic data. More specifically, in the case of existing systems that attempt to facilitate the integration of external diagnostic applications, these systems often compromise the security of sequencing data (and other information) in exchange for the flexibility that allows external workflows (including internal and/or external applications) to access sequencing data for diagnostic purposes. Indeed, some conventional systems are vulnerable to harmful external diagnostic workflows that maliciously or unintentionally damage or corrupt sequencing data related to nucleotide-based calls and/or sequencing data associated with other diagnostic applications. As a result, many of these existing diagnostic systems fail to meet one or more diagnostic criteria required by various regulatory bodies (e.g., in vitro diagnostic criteria set by the U.S. Food and Drug Administration). Additionally, some conventional diagnostic systems are inefficient. More specifically, existing systems often consume computing resources inefficiently when performing genomic diagnostic workflows. For example, some existing systems lack internal programming or other considerations for computing resource management, instead flooding available processors and memory with numerous (simultaneous) requests and instructions for the many processes involved in generating diagnostic data (often resulting in backlogs and slowdowns). Therefore, such existing systems are slow to produce the requested diagnostic results for the diagnostic workflow, or are unable to produce results completely, instead resulting in computational errors. This disclosure describes a method, a non-temporary computer-readable medium, and embodiments of a system that can flexibly, securely, and efficiently facilitate the execution of external workflows for the diagnostic analysis of nucleotide sequencing data. For example, the disclosed system may utilize a container orchestration engine to enable an external system (e.g., a third-party system) to generate and implement a workflow for analyzing sequencing data (e.g., sequencing data generated by a sequencing instrument and/or variant analysis model). In some cases, the disclosed system further utilizes the container orchestration engine to generate sequencing data, such as nucleotide base calls, in order to implement a diagnostic workflow for analyzing the sequencing data. For example, the disclosed system may utilize the container orchestration engine to identify workflow containers that compartmentally define the individual functions of the workflow. In some such cases, the disclosed system may require the container orchestration engine to perform an external sequencing diagnostic workflow outside of the workflow for the variant analysis model and to implement the external sequencing diagnostic workflow in the workflow container. When executing individual workflow containers, the d