DCT

1:21-cv-01635

Invitae Corp v. Natera Inc

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 1:21-cv-01635, D. Del., 11/21/2021
  • Venue Allegations: Venue is alleged to be proper in the District of Delaware because the Defendant, Natera, Inc., is a Delaware corporation.
  • Core Dispute: Plaintiff alleges that Defendant’s Signatera test for minimal residual disease infringes two patents related to methods for assembling DNA sequence data to accurately identify genetic mutations.
  • Technical Context: The dispute is in the field of bioinformatics for next-generation sequencing (NGS), where computational methods are used to piece together millions of short DNA "reads" to reconstruct a larger genetic sequence and detect variations from a reference.
  • Key Procedural History: The complaint makes specific reference to the prosecution history of the patents-in-suit, quoting from the U.S. Patent and Trademark Office’s Notices of Allowance to argue that the claimed inventions are not directed to a patent-ineligible abstract idea and were not routine or conventional at the time of invention.

Case Timeline

Date Event
2012-04-04 Earliest Priority Date for ’308 and ’863 Patents
2017-08-01 Alleged commercial launch of Natera's Signatera test
2019-05-01 Alleged launch of Signatera test for clinical use
2021-10-19 U.S. Patent No. 11,149,308 Issues
2021-10-26 U.S. Patent No. 11,155,863 Issues
2021-11-21 Complaint Filed

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

U.S. Patent No. 11,149,308, "Sequence Assembly" (issued October 19, 2021)

The Invention Explained

  • Problem Addressed: The patent describes prior art methods for analyzing DNA sequence information from next-generation sequencing (NGS) as "problematic" (Compl. ¶14). Specifically, aligning very short DNA "reads" directly to a long reference genome offered "little positional accuracy" and made it "difficult or impossible to detect" certain mutations like insertions or deletions ("indels") (’308 Patent, col. 2:10-17). Furthermore, existing alignment algorithms faced an undesirable "trade-off between substitution sensitivity and deletion sensitivity," meaning that tuning an algorithm to find one type of mutation well would likely cause it to miss the other (’308 Patent, col. 1:65–2:3).
  • The Patented Solution: The patent teaches a multi-stage computational method to improve accuracy. First, the numerous short sequence reads are assembled into a longer, intermediate sequence called a "contig." This contig is then aligned to a reference genome. Finally, the original short reads are aligned to the now-positioned contig. By combining the information from both the contig-to-reference alignment and the read-to-contig alignments, the method can provide a more accurate and complete description of any mutations present in the original DNA sample relative to the reference genome (’308 Patent, col. 4:41-53).
  • Technical Importance: This two-stage alignment approach claims to overcome the sensitivity trade-offs of prior art methods, enabling the identification of mutation patterns that were previously difficult to detect with fidelity, such as "long indels, indel-proximal substitutions, and indels near the ends of reads" (’308 Patent, col. 4:48-51).

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶26-27).
  • The essential elements of Claim 1 include:
    • A method for identifying differences between a reference human genome and sequence reads from a human subject.
    • Obtaining nucleic acid from the subject and using next-generation sequencing to generate sequence reads.
    • Genotyping the reads using a "multi-stage alignment" performed on a computer, which comprises the steps of:
      • Assembling a "contig" from the sequence reads.
      • Aligning the contig to the reference genome using a "first substitution probability and a first gap penalty" to find "first differences."
      • Aligning the sequence reads to the contig using a "second substitution probability and a second gap penalty" to find "second differences."
      • Identifying multiple mutations by "combining the reference alignment and the sequence read alignments" based on the first and second differences.
  • The complaint does not explicitly reserve the right to assert other claims, but its infringement allegations are not limited to claim 1 (Compl. ¶33).

U.S. Patent No. 11,155,863, "Sequence Assembly" (issued October 26, 2021)

The Invention Explained

  • Problem Addressed: The ’863 patent addresses the same technical problems as the ’308 patent, citing identical passages from its specification regarding the low positional accuracy and indel detection difficulties of prior art methods (Compl. ¶14; ’863 Patent, col. 2:10-17). It also notes the same trade-off between mismatch and indel sensitivity (Compl. ¶15; ’863 Patent, col. 1:65–2:3).
  • The Patented Solution: The ’863 patent also discloses a multi-stage alignment solution where reads are first assembled into a contig, the contig is aligned to a reference, and the reads are then aligned back to the contig (’863 Patent, col. 4:41-53). The solution is conceptually identical to that of the ’308 patent, with variations in the specific claim language used to describe the process.
  • Technical Importance: The claimed technique is described as enabling analyses that were "previously computationally intractable" and overcoming the limitations of prior art alignment trade-offs, consistent with the ’308 patent (Compl. ¶23; ’863 Patent, col. 4:51-53).

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶28-29).
  • The essential elements of Claim 1 include:
    • A method for assembling and aligning sequence reads having mutations.
    • Sequencing a sample to generate reads, where the sequencing comprises fragmenting nucleic acid, attaching fragments to a flow cell, and amplifying them to create clusters.
    • Using a computer system to perform the steps of:
      • Assembling a "contig" from the sequence reads.
      • Identifying "contig-to-reference descriptions" by aligning the contig to the reference genome.
      • Identifying "read-to-contig descriptions" by aligning the sequence reads to the contig.
      • Generating a "read-to-reference description" by "aligning" the contig-to-reference descriptions with the read-to-contig descriptions to map positional information of the mutations.
  • The complaint's infringement allegations are not limited to claim 1 (Compl. ¶38).

III. The Accused Instrumentality

Product Identification

The accused instrumentality is Defendant Natera’s "Signatera" test (Compl. ¶5).

Functionality and Market Context

The Signatera test is described as a non-invasive, commercial liquid biopsy test for cancer diagnostics, used to detect molecular residual disease (Compl. ¶5, 24, 25). The complaint alleges that the technology underlying the test involves collecting patient tissue and blood samples, performing whole exome sequencing to compare tumor and normal DNA, and thereby identifying somatic mutations (Compl. ¶25, 26). Based on this analysis, a custom monitoring panel is generated (Compl. ¶25). The complaint alleges, on information and belief, that Natera performs this test at its CLIA laboratory in California and that the mutation identification process uses the "Genome Analysis Toolkit ('GATK') HaplotypeCaller or a method that implements a similar analysis, such as Mutect2" (Compl. ¶5, 24, 26, 28).

No probative visual evidence provided in complaint.

IV. Analysis of Infringement Allegations

The complaint references preliminary claim charts in Exhibits 7 and 8, but these exhibits were not provided with the complaint document. The narrative infringement theory is summarized below.

  • ’308 Patent Infringement Allegations: The complaint alleges that when Natera performs the Signatera test, it practices the multi-stage alignment method of claim 1 (Compl. ¶26). The core of the allegation is that Natera's use of bioinformatics tools, such as GATK HaplotypeCaller or Mutect2, to identify somatic mutations constitutes infringement of the claimed steps (Compl. ¶26). Plaintiff alleges that this process involves the claimed sequence of assembling reads into a contig, aligning the contig to a reference, aligning reads back to the contig, and combining the results to genotype mutations (Compl. ¶16, 26).
  • ’863 Patent Infringement Allegations: The infringement theory for the ’863 Patent is substantially the same, alleging that Natera's Signatera test, through its use of tools like GATK HaplotypeCaller or Mutect2, performs the steps recited in claim 1 (Compl. ¶28). This includes the generation of sequence reads via a flow cell, the assembly of a contig, the creation of contig-to-reference and read-to-contig "descriptions," and the combination of these descriptions to map mutations (Compl. ¶17, 28).
  • Identified Points of Contention:
    • Technical Questions: A central technical question will be whether the actual workflow of the accused bioinformatics software (e.g., GATK HaplotypeCaller) maps onto the specific steps recited in the claims. For example, what evidence demonstrates that the accused process uses two distinct sets of alignment parameters ("first substitution probability and a first gap penalty" vs. "a second...") as required by claim 1 of the ’308 Patent? For the ’863 Patent, what technical steps in the accused process correspond to "identifying... descriptions" and then "aligning" those descriptions?
    • Scope Questions: The infringement analysis may turn on the scope of key terms. For instance, does the term "contig," as used in the patents, read on the intermediate data structures (e.g., locally re-assembled haplotypes) generated by software like GATK HaplotypeCaller? The degree of overlap between the patent's specific multi-stage process and the operational details of Natera's accused test will be a primary focus.

V. Key Claim Terms for Construction

  • The Term: "contig" (asserted in claim 1 of both patents)

    • Context and Importance: This term is foundational to the claimed multi-stage process, which requires assembling reads into this intermediate structure before aligning to a reference. Practitioners may focus on this term because the accused software may not use the term "contig" to describe its intermediate data structures; the dispute will be whether those structures are functionally and structurally equivalent to what the patent discloses as a contig.
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: The specification provides a functional definition, stating a contig can be a "consensus sequence, i.e., a probable interpretation of the sequence of the nucleic acid represented by that contig" (’308 Patent, col. 4:50-53).
      • Evidence for a Narrower Interpretation: The detailed description discusses specific methods for creating contigs, such as "overlap consensus assembly" or "de Bruijn graph[s]" (’308 Patent, col. 13:30-36). This could support an argument that the term is limited to structures created by such explicit de novo or reference-guided assembly methods.

  • The Term: "multi-stage alignment" (’308 Patent, claim 1)

    • Context and Importance: This term encapsulates the entire inventive concept of the ’308 Patent. Infringement depends on whether Natera's process performs an alignment that meets this definition.
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: A party might argue the term broadly covers any bioinformatics process that uses more than one distinct alignment phase to improve accuracy over a single direct alignment.
      • Evidence for a Narrower Interpretation: The body of claim 1 itself defines what the "multi-stage alignment" comprises, explicitly listing the sequence of: (1) assembling a contig, (2) aligning the contig to a reference, (3) aligning reads to the contig, and (4) genotyping by combining the alignments. A party could argue that the term is strictly limited to this recited sequence of steps.

VI. Other Allegations

  • Indirect Infringement: The complaint does not plead any counts of indirect infringement (induced or contributory). The infringement counts are for direct infringement under 35 U.S.C. § 271(a) (Compl. ¶33, 38).
  • Willful Infringement: The complaint does not contain an explicit allegation of willful infringement. The patents-in-suit issued in late October 2021, and the complaint was filed in November 2021, making pre-suit knowledge highly improbable. The prayer for relief requests a finding that the case is "exceptional" under 35 U.S.C. § 285, but no specific facts supporting willfulness are alleged in the body of the complaint (Compl. p. 16).

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

  • A core issue will be one of evidentiary mapping: can Plaintiff produce evidence to demonstrate that the internal, proprietary workings of Defendant's Signatera test—allegedly using tools like GATK HaplotypeCaller—perform the specific, ordered steps of the "multi-stage alignment" as claimed? This will involve a technical deep-dive into how the accused software operates and how that operation maps to claim limitations such as "assembling a contig" and using distinct alignment parameters.
  • A second central question will be one of patent eligibility: although the complaint preemptively cites favorable prosecution history, will the asserted method claims, which recite computational steps for analyzing genetic data, be found to be directed to a patent-ineligible abstract idea or law of nature under 35 U.S.C. § 101? The court will have to determine if the claims are merely an abstract process of data analysis or a concrete improvement to computer functionality in the field of DNA sequencing.