DCT

1:24-cv-00680

Molecular Loop Biosciences Inc v. Illumina Inc

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 1:24-cv-00680, D. Del., 06/10/2024
  • Venue Allegations: Venue is alleged to be proper in the District of Delaware because both Defendants are Delaware corporations that reside and conduct business in the state.
  • Core Dispute: Plaintiff alleges that Defendant’s Next Generation Sequencing (NGS) products and services, including the Verifi, VeriSeq, and TruSight lines of genetic tests, infringe three patents related to methods for improving the accuracy of nucleic acid sequencing.
  • Technical Context: The technology addresses errors inherent in high-throughput DNA sequencing by using unique molecular tags or barcodes to identify original DNA molecules, allowing for the detection and correction of artifacts introduced during sample preparation and amplification.
  • Key Procedural History: The complaint alleges that the ’200 Patent is a continuation of the ’852 Patent. It also alleges Defendants were aware of the technology and patents since at least February 2023 due to discussions between the parties, and became aware of infringement through communications in December 2023.

Case Timeline

Date Event
2009-04-30 U.S. Patent No. 11,840,730 Priority Date
2010-12-23 U.S. Patent No. 11,041,852 Priority Date
2010-12-23 U.S. Patent No. 11,768,200 Priority Date
2012-03-XX Defendant VHI launches Accused Verifi Prenatal Test
2013-01-07 Defendant Illumina announces acquisition of VHI
2017-XX-XX Defendants launch Accused Verifi Prenatal Plus Test
2017-04-XX Defendant Illumina launches Accused VeriSeq NIPT solution
2018-10-30 Defendant Illumina launches Accused TruSight Oncology 500 test
2019-06-XX Defendants launch Accused VeriSeq NIPT Solution v2
2019-11-05 Defendant Illumina launches Accused TruSight Oncology 500 ctDNA and HT tests
2021-06-22 U.S. Patent No. 11,041,852 Issued
2023-09-26 U.S. Patent No. 11,768,200 Issued
2023-12-12 U.S. Patent No. 11,840,730 Issued
2024-06-10 Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 11,041,852 - “Methods for Maintaining the Integrity and Identification of a Nucleic Acid Template in a Multiplex Sequencing Reaction”

The Invention Explained

  • Problem Addressed: In multiplex sequencing, where samples from different sources are pooled, template DNA molecules can physically overlap during amplification on a solid surface (e.g., a flow cell). This can create "chimeric" molecules that combine a barcode from one sample with a gene sequence from another, leading to incorrect assignment of sequencing data. (Compl. ¶19; ’852 Patent, col. 1:35-55).
  • The Patented Solution: The invention proposes incorporating at least two unique identifier sequences (a "distinct pair" of barcodes) into each DNA template, flanking the sequence of interest. If a chimeric molecule forms due to crossover, it will possess an incorrect combination of identifiers (e.g., barcode 1 from sample A and barcode 2 from sample B). This invalid pairing allows the resulting sequence read to be identified as an error and discarded during data analysis, thereby reducing crossover error. (Compl. ¶30; ’852 Patent, col. 6:55-7:38). Figure 2 of the patent, reproduced in the complaint, illustrates how a chimeric molecule formed during amplification (panel 3) results in a sequence with an invalid barcode pair (A1/B2), allowing for its identification and removal (panel 4). (Compl. ¶31, p. 9).
  • Technical Importance: This dual-identifier error-checking method provided a way to improve the accuracy and reliability of high-throughput, multiplexed NGS analyses by actively identifying and removing specific types of artifacts generated during sample preparation. (Compl. ¶¶22-23).

Key Claims at a Glance

  • The complaint asserts infringement of at least independent Claim 1. (Compl. ¶¶28, 86).
  • Claim 1 of the ’852 Patent requires a method with the following essential elements:
    • incorporating at least two members of a plurality of identifier sequences into template nucleic acids obtained from at least two different samples, wherein said two members constitute a distinct pair associated with said template;
    • combining said nucleic acid templates into a single sample;
    • amplifying said nucleic acid templates on a surface of a flow cell thereby to form clusters, at least one of which comprises a chimeric sequence comprising a combination of identifier sequences that are different than any of said distinct pairs;
    • sequencing amplicons obtained in said amplifying step; and
    • discarding sequence reads obtained in said sequencing step that contain said combination of identifier sequences that are different than any of said distinct pairs, thereby to reduce cross-over error.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

U.S. Patent No. 11,768,200 - “Methods for Maintaining the Integrity and Identification of a Nucleic Acid Template in a Multiplex Sequencing Reaction”

The Invention Explained

  • Problem Addressed: As a continuation of the ’852 Patent, the ’200 Patent addresses the same problem of errors introduced during NGS workflows that compromise the ability to accurately analyze sequencing reads, particularly in multiplex applications. (Compl. ¶¶19-22).
  • The Patented Solution: The invention describes a method for validating the sequence of a nucleic acid analyte by using two or more identifier sequences. A first identifier is incorporated into the 5' portion of the analyte and a second is incorporated into the 3' portion. The validation step involves sequencing the identifiers and the analyte, and then "analyzing both identifiers and excluding the sequences" that contain only one identifier or an "incorrect pair of identifiers," thereby ensuring the integrity of the data. (Compl. ¶36; ’200 Patent, Claim 1).
  • Technical Importance: This method provides a framework for using paired identifiers not just to detect errors, but as a specific validation step to filter and correct the final sequencing data, allowing sequences to be confidently traced back to their original sample. (Compl. ¶37; ’200 Patent, col. 9:5-11).

Key Claims at a Glance

  • The complaint asserts infringement of at least independent Claim 1. (Compl. ¶¶36, 103).
  • Claim 1 of the ’200 Patent requires a method with the following essential elements:
    • detecting the presence of two or more identifier sequences that are uniquely associated with the nucleic acid analyte of interest, wherein at least a first identifier sequence is incorporated into a 5' portion... and at least a second... is incorporated into a 3' portion... and wherein the first and second identifier sequences have four or more nucleotides and are different;
    • sequencing the first identifier sequence, the nucleic acid analyte of interest and the second identifier sequence; and
    • validating the sequence of the nucleic acid analyte of interest by analyzing both identifiers and excluding the sequences of those nucleic acid analytes of interest containing only one identifier or an incorrect pair of identifiers from sequence analysis.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

U.S. Patent No. 11,840,730 - “Methods and Compositions for Evaluating Genetic Markers”

Multi-Patent Capsule: U.S. Patent No. 11,840,730

  • Patent Identification: U.S. Patent No. 11,840,730, “Methods and Compositions for Evaluating Genetic Markers,” issued December 12, 2023. (Compl. ¶38).
  • Technology Synopsis: The patent addresses bias introduced during nucleic acid amplification, such as the over- or under-representation of certain starting molecules, which can lead to incorrect genotype calls. (Compl. ¶¶40-41; ’730 Patent, col. 4:57-5:11). The solution involves tagging individual nucleic acid molecules with unique "differentiator tags" (also known as UMIs) before amplification. After sequencing, all reads that share the same target sequence and differentiator tag are assumed to have originated from a single starting molecule and are "collapsing" into a single count, thereby correcting for amplification bias. (Compl. ¶40; ’730 Patent, Abstract). Figure 5 of the patent, shown in the complaint, illustrates how this collapsing step corrects an erroneous "homozygous T" call to a correct "heterozygous A/T call." (Compl. ¶20, p. 6).
  • Asserted Claims: At least independent Claim 1 is asserted. (Compl. ¶¶40, 120).
  • Accused Features: The complaint accuses the TruSight Oncology (TSO) 500 Tests, which allegedly use UMIs to reduce error rates and employ software that performs error correction by "collapsing the sequence" to remove duplicate reads and sequencing errors. (Compl. ¶¶67, 69, 120).

III. The Accused Instrumentality

Product Identification

  • The accused instrumentalities are Defendants’ genetic testing products and services, grouped into three main families: the Verifi Prenatal Tests, the VeriSeq NIPT Tests, and the TruSight Oncology (TSO) 500 Tests. (Compl. ¶¶5-8).

Functionality and Market Context

  • The complaint alleges that all accused products are methods that utilize Next Generation Sequencing (NGS) and incorporate identifier sequences to reduce errors. (Compl. ¶¶5-9).
  • Verifi Prenatal Tests: These are non-invasive prenatal tests that analyze cell-free DNA from maternal blood to screen for chromosomal aneuploidies. The workflow allegedly requires preparing sequencing libraries with Illumina TruSeq kits, which incorporate at least two identifier sequences, pooling up to 96 samples in a single flow cell, amplifying the DNA on Illumina sequencers, and using software to demultiplex samples and discard reads without a matching index pair. (Compl. ¶¶47-52). The complaint provides a workflow diagram showing the steps of the Verifi test, including Library Preparation, Sequencing, and Data Analysis. (Compl. p. 18, Fig. 1).
  • VeriSeq NIPT Tests: These are in vitro diagnostic screening tests for genome-wide anomalies, also using cell-free DNA from maternal blood. The process is alleged to require using unique "dual index adapters" for each sample during library preparation, pooling samples, and using Illumina sequencers that capture and amplify libraries on the solid surface of a flow cell. (Compl. ¶¶54-57).
  • TSO 500 Tests: These are oncology assays designed to analyze biomarkers in tissue or liquid biopsies. The workflow allegedly requires a library preparation step that ligates "unique molecular identifiers (UMIs)" to barcode each individual DNA strand. The workflow further allegedly includes software (e.g., DRAGEN) that performs error correction by collapsing sequence reads with shared UMIs down to unique reads to filter false positives. (Compl. ¶¶64, 67, 69). The complaint includes a workflow diagram for the TSO 500 tests, which explicitly shows a "Library prep" step that introduces identifiers, followed by "Sequencing," "Variant calling," and "Insights and reporting." (Compl. ¶65, p. 24).

IV. Analysis of Infringement Allegations

’852 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
incorporating at least two members of a plurality of identifier sequences into template nucleic acids... wherein said two members constitute a distinct pair associated with said template The Verifi Tests allegedly use Illumina TruSeq library preparation kits to incorporate at least two identifier sequences (dual indexes) into template DNA. ¶86a col. 6:55-58
combining said nucleic acid templates into a single sample The accused workflow allegedly involves pooling up to 96 uniquely indexed samples together for sequencing in a single flow cell lane. ¶86b col. 5:1-4
amplifying said nucleic acid templates on a surface of a flow cell thereby to form clusters... The Verifi Tests are allegedly run on Illumina sequencers (e.g., HiSeq 2000) that use flow cell technology to form clusters via DNA amplification. ¶86c col. 10:49-55
sequencing amplicons obtained in said amplifying step The accused workflow uses Illumina HiSeq and NextSeq systems to perform NGS sequencing of the amplified DNA. ¶86d col. 1:24-34
discarding sequence reads... that contain said combination of identifier sequences that are different than any of said distinct pairs, thereby to reduce cross-over error The Verifi Tests allegedly use Illumina software (bcl2fastq2) to demultiplex samples, which places reads without a matching index adapter pair into an "Undetermined" file, thereby discarding them from the main analysis. ¶86e col. 7:35-38
  • Identified Points of Contention:
    • Functional Questions: A central question may be whether the accused software's function of sorting reads that lack a valid index pair into an "Undetermined" file performs the specific function claimed. Does this process affirmatively identify and discard reads containing a "combination of identifier sequences that are different than any of said distinct pairs," as recited in the claim, or does it merely filter for reads that fail to match a pre-defined list of expected pairs?
    • Scope Questions: The dispute may turn on the construction of "discarding." Does placing a read in a separate file for non-matching indexes, which may or may not be used for other purposes, constitute "discarding" in the context of the claim, which links the act to the purpose of reducing "cross-over error"?

’200 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
detecting the presence of two or more identifier sequences... wherein at least a first identifier sequence is incorporated into a 5' portion... and... a second... into a 3' portion... and wherein the first and second... have four or more nucleotides and are different The Verifi Tests allegedly use Illumina TruSeq kits to incorporate dual indexes (first and second identifier sequences, each with more than four nucleotides) into the 5' and 3' portions of template DNA. ¶103a col. 9:14-28
sequencing the first identifier sequence, the nucleic acid analyte of interest and the second identifier sequence The Verifi Tests are allegedly performed on Illumina sequencers that provide the throughput and read length to sequence the indexes and the DNA sample. ¶103b col. 1:24-28
validating the sequence... by analyzing both identifiers and excluding the sequences... containing only one identifier or an incorrect pair of identifiers from sequence analysis The accused workflow allegedly uses software to demultiplex the data, which involves analyzing the indexes and excluding reads with incorrect indexes (placing them in an "Undetermined" file) from the final analysis of the sample pool. ¶103c col. 12:28-34
  • Identified Points of Contention:
    • Functional Questions: Similar to the '852 patent, a key point of contention may be whether the accused demultiplexing software performs the claimed step of "validating... by... excluding... sequences... containing... an incorrect pair of identifiers." The defense may argue that simply failing to find a match on a sample sheet is technically different from affirmatively identifying and excluding an "incorrect pair."
    • Scope Questions: What constitutes an "incorrect pair of identifiers"? Does this term require a specific chimeric combination (e.g., index 1 from Sample A with index 2 from Sample B), or does it broadly cover any index pair that is not on the list of expected valid pairs for that sequencing run?

V. Key Claim Terms for Construction

  • The Term: "discarding sequence reads... that contain said combination of identifier sequences that are different than any of said distinct pairs" (’852 Patent, Claim 1)
  • Context and Importance: This term defines the active error-correction step of the invention. Its construction will be critical to determining whether standard demultiplexing software, which sorts reads based on matching indexes from a sample sheet, infringes. Practitioners may focus on this term because the plaintiff's theory appears to equate sorting non-matching reads into a separate file with the affirmative act of "discarding" reads containing a newly formed, invalid combination.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification states that after a chimeric molecule with an incorrect pair of identifiers is formed, "data analysis could recognize that the molecule contains an invalid identifier pair" and "reads for these molecules could be identified as error and discarded in data analysis." (’852 Patent, col. 7:35-38). This general language could support a broad functional definition of "discarding."
    • Evidence for a Narrower Interpretation: The patent's background and detailed description focus on solving the specific problem of "template cross-over" that creates "chimeric" sequences. (Compl. ¶¶19, 28; ’852 Patent, col. 6:55-7:2). A narrower interpretation might require that the "discarding" step be specifically directed at identifying and removing these crossover-generated combinations, rather than just filtering out any read with an unrecognized index pair, which could arise from other error types.
  • The Term: "validating the sequence... by analyzing both identifiers and excluding the sequences... containing... an incorrect pair of identifiers" (’200 Patent, Claim 1)
  • Context and Importance: This term is the core of the validation method in the ’200 Patent. The infringement analysis will likely depend on whether the accused demultiplexing process meets this functional limitation. The dispute may hinge on whether sorting reads constitutes "validating by excluding."
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification teaches that the "invention is based upon using two or more identifiers... that are uniquely associated with an analyte" and that "template containing... an incorrect pair... will be excluded from analysis." (’200 Patent, col. 2:15-26). This supports a broad reading where any exclusion based on an incorrect pair serves the validation function.
    • Evidence for a Narrower Interpretation: The patent explains that the use of a pair of identifiers "allows a sequence to be traced back to that sample during analysis." (’200 Patent, col. 9:5-28). A narrower reading could suggest that "validating" requires more than simple filtering; it might imply a positive confirmation of valid pairs and an explicit exclusion of specific "incorrect pairs" to ensure data integrity, a potentially higher standard than simply sorting unmatched reads.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges Defendants induce infringement by selling and supplying the accused tests to third-party laboratories with the knowledge and intent that they will be used in an infringing manner. This is allegedly supported by providing instructions, manuals, software, and technical support that direct users to perform the claimed methods. (Compl. ¶¶88-89, 105-106, 122-123).
  • Willful Infringement: Willfulness is alleged based on Defendants' knowledge of the patents since at least February 2023, stemming from discussions and communications with the Plaintiff, and from Defendants' own research into related patent applications. The complaint alleges that despite this knowledge, Defendants continued to commercialize the accused products without a license. (Compl. ¶¶79-82, 97, 114, 126).

VII. Analyst’s Conclusion: Key Questions for the Case

  • A core issue will be one of functional operation: Does the standard demultiplexing process in Illumina’s software, which categorizes reads into "matched" and "undetermined" files based on a sample sheet, perform the specific, affirmative step of "discarding" or "excluding" reads because they contain a new, chimeric "combination of identifier sequences that are different" from the valid pairs, as required by the claims? Or is there a fundamental mismatch between this filtering process and the claimed error-correction mechanism?
  • A related question will be one of definitional scope: Can the claim terms "incorrect pair of identifiers" and "combination of identifier sequences that are different than any of said distinct pairs" be construed to cover any read whose indexes simply do not match an expected list, or do they require the identification of a specific type of error product, such as a molecule formed by a template crossover event?
  • For the ’730 patent, a key evidentiary question will be one of technical equivalence: Does the accused TSO 500 software's process of using UMIs to "collapse the sequence" and "remove duplicate reads" perform the same function in substantially the same way to achieve the same result as the claimed method of "collapsing target:differentiator tag combinations... into a single count" for the purpose of correcting for amplification bias?