Derivative works — US Patent 11566276

Defensive Disclosure and Prior Art Generation for Analyte Detection Technologies

Publication Date: April 26, 2026
Reference Technology: The core inventive concept described in US Patent 11,566,276 B2 ("Compositions and methods for analyte detection"), which involves the use of probe-based detection reagents conjugated to unique, pre-assigned nucleic acid labels for in-situ multiplexed analyte detection via amplification and sequencing of the labels.
Purpose: This document is intended to enter the public domain as prior art. It discloses a series of derivative works, improvements, and alternative embodiments of the reference technology. The descriptions provided herein are enabling for a Person Having Ordinary Skill in the Art (PHOSITA).

Derivative Embodiments Based on Core Method Claims (Ref: Claim 1)

Axis 1: Material & Component Substitution

1.1. Cryo-Fixation with Non-Enzymatic, Light-Activated Amplification

Enabling Description: This method bypasses chemical fixation (e.g., paraformaldehyde) to better preserve antigenicity and nucleic acid integrity. A biological sample is prepared by flash-freezing in liquid nitrogen, followed by freeze-substitution with a non-aqueous solvent (e.g., acetone) containing the detection reagents. After probe hybridization at low temperatures, the sample is embedded in a transparent, cryo-compatible matrix (e.g., a glycerol-sucrose mixture). The nucleic acid labels are designed to incorporate photocaged nucleotides. Amplification is achieved via a non-enzymatic, photo-initiated chemical ligation chain reaction. The sample is illuminated with a specific wavelength of light to de-protect the caged nucleotides, allowing templated ligation to occur. Sequential cycles of ligation, washing, and imaging are performed to sequence the labels in situ. This method avoids heat-denaturation steps, preserving the tissue architecture with sub-micron resolution.

Mermaid Diagram:

graph TD
    A[Start: Biological Sample] --> B{Flash-Freezing};
    B --> C[Freeze-Substitution with Detection Reagents in Acetone];
    C --> D[Embed in Cryo-Compatible Matrix];
    D --> E{Hybridization & Wash};
    E --> F[Photo-Initiated Ligation Chain Reaction];
    F --> G{Cyclical Imaging & Ligation};
    G --> H[Sequence Assembly & Spatial Mapping];
    H --> I[End: Analyte Map];

1.2. Mass Spectrometry Readout via Photocleavable Mass-Tags

Enabling Description: In this variation, the sequencing step is replaced with mass spectrometry for readout. Each unique nucleic acid label is conjugated not only to the probe but also to a unique, photocleavable mass-tag (e.g., from the MeCAT kit family) of a pre-determined mass. After in situ hybridization and washing, a high-resolution laser is used to scan the sample. At each pixel (x,y), the laser cleaves the mass-tags from the hybridized labels. The liberated ions are collected and analyzed by a coupled time-of-flight (TOF) mass spectrometer. The resulting mass spectrum at each pixel reveals the identity of the analytes present at that location. This approach eliminates the need for amplification and complex sequencing chemistry.

Mermaid Diagram:

sequenceDiagram
    participant Laser as Scanning Laser
    participant Sample as Sample Surface
    participant MS as Mass Spectrometer
    Laser->>Sample: Irradiate pixel (x,y)
    Sample->>MS: Release photocleaved mass-tags
    MS->>MS: Analyze mass spectrum
    MS-->>Laser: Record masses at (x,y)
    Note over Laser,MS: Repeat for all pixels to build image

Axis 2: Operational Parameter Expansion

2.1. High-Pressure/High-Temperature Detection in Deep-Sea Vent Biofilms

Enabling Description: This method is adapted for analyzing microbial life in extreme environments. A biofilm sample is collected from a deep-sea hydrothermal vent using a robotic submersible and immediately fixed in situ within a pressure-retaining chamber using a glutaraldehyde-based fixative tolerant to high temperatures. The detection reagents use probes (e.g., aptamers selected for thermostability) conjugated to nucleic acid labels synthesized with locked nucleic acids (LNAs) to increase the melting temperature (Tm) and stability against thermal degradation. All hybridization, washing, and amplification steps are carried out within the high-pressure (e.g., >200 atm), high-temperature (e.g., 60-90°C) chamber using a thermostable DNA polymerase (e.g., PfuUltra II). The sequencing readout is performed after depressurization or via a pressure-compatible microfluidic sequencing cell.

Mermaid Diagram:

stateDiagram-v2
    [*] --> PressurizedFixation
    PressurizedFixation --> HighTmHybridization: Sample Fixed In-Situ
    HighTmHybridization --> Wash: Probes Bound
    Wash --> ThermostableAmplification: Background Removed
    ThermostableAmplification --> Sequencing: Labels Amplified
    Sequencing --> [*]: Data Acquired
    state HighTmHybridization {
        note right of HighTmHybridization : Probes use LNA-modified labels for stability at >80°C
    }

Axis 3: Cross-Domain Application

3.1. Aerospace: Spatially-Resolved Material Fatigue Analysis

Enabling Description: This method is repurposed to create a spatial map of micro-cracks and metallic strain in aerospace components (e.g., a turbine blade). The detection probes are not antibodies but are instead engineered peptides or small molecules that selectively bind to newly exposed metallic crystal faces or oxides characteristic of material fatigue. Each unique probe is conjugated to a DNA barcode. The component surface is incubated with a pool of these detection reagents. After washing, the surface is coated with a thin, transparent polymer matrix to fix the reagents' locations. The DNA barcodes are then amplified and sequenced in situ using a portable sequencing device. The resulting data provides a high-resolution map of fatigue hotspots, predicting potential failure points long before they are visible through conventional imaging.

Mermaid Diagram:

flowchart LR
    subgraph TurbineBlade [Turbine Blade Surface]
        A[Micro-crack]
        B[Strain Zone]
    end
    subgraph Reagents [Detection Reagents]
        C(Probe: Fatigue-Marker Peptide)
        D(Label: Unique DNA Barcode)
        C---D
    end
    Reagents --> TurbineBlade
    TurbineBlade --> E{Coating & Fixation}
    E --> F[In-Situ Sequencing]
    F --> G((Map of Fatigue Hotspots))

3.2. AgTech: Pathogen and Nutrient Mapping in Soil Micropores

Enabling Description: To analyze soil health, a soil core sample is embedded in a clear, water-permeable resin (e.g., LR White acrylic resin) and sectioned. The section is incubated with a reagent library where probes target specific bacterial 16S rRNA sequences (for pathogen identification), fungal ITS sequences, and chelated ions (e.g., phosphate, nitrate) using ion-specific aptamers. Each probe is barcoded. Following hybridization, the soil matrix pores are filled with an amplification solution, and thermal cycling is performed on the entire slide. In-situ sequencing reveals a micron-scale map showing the co-localization of specific pathogens with nutrient gradients, allowing for precision application of fertilizers or bacteriophages.

Mermaid Diagram:

erDiagram
    SOIL_MICROPORE {
        int x_coord
        int y_coord
        int z_coord
    }
    ANALYTE {
        string type
        string name
    }
    PROBE {
        string sequence
        string target
    }
    SOIL_MICROPORE ||--o{ ANALYTE : contains
    ANALYTE ||--|{ PROBE : is_detected_by

3.3. Consumer Electronics: Wafer Contamination Quality Control

Enabling Description: For semiconductor quality control, this method maps organic and metallic contaminants on a silicon wafer. Detection reagents consist of aptamers or nanobodies selected to bind with high specificity to common contaminants (e.g., sodium ions, iron, specific photoresist residues). These probes are tagged with DNA barcodes. The wafer is scanned by a microfluidic head that dispenses the reagent library, washes, and then performs the amplification and sequencing chemistry directly on the wafer surface. The output is a digital map of the wafer, highlighting contaminant locations with parts-per-billion sensitivity, allowing fabrication lines to identify the source of the contamination (e.g., a specific chemical bath or handling robot).

Mermaid Diagram:

gantt
    title Wafer Contamination Mapping Workflow
    dateFormat  YYYY-MM-DD-HH
    section Wafer Processing
    Dispense Reagents  :a1, 2026-04-26-09, 10m
    Hybridization      :a2, after a1, 30m
    Wash Step 1        :a3, after a2, 5m
    section On-Wafer Analysis
    Amplification      :b1, after a3, 60m
    Sequencing Cycles  :b2, after b1, 120m
    section Data Output
    Generate Contaminant Map :c1, after b2, 15m

Axis 4: Integration with Emerging Tech

4.1. AI-Optimized Barcode Design and Spatial Deconvolution

Enabling Description: The design of the unique nucleic acid labels is outsourced to an AI model. A generative adversarial network (GAN) is trained on existing genomic and transcriptomic data from relevant species to generate a library of thousands of barcode sequences that are optimized for (1) minimal off-target hybridization, (2) uniform amplification efficiency, and (3) maximal error correction capability (e.g., high Levenshtein distance). For data analysis, a graph neural network (GNN) is employed. The GNN treats each detected molecule as a node in a graph, with edges representing spatial proximity. The GNN then learns to classify cells and identify complex tissue microenvironments based on the spatial arrangement of dozens or hundreds of different analyte types, revealing biological patterns not discernible by human analysis.

Mermaid Diagram:

flowchart TD
    A[Genomic Data] --> B[GAN Barcode Generator]
    B --> C{Optimized Barcode Library};
    subgraph Experiment
        D[In-Situ Sequencing] --> E[Raw Spatial Data (x,y,z,seq)];
    end
    C --> D
    E --> F[Graph Neural Network];
    F --> G[Classified Cells & Tissue Maps];

Axis 5: The "Inverse" or Failure Mode

5.1. Limited-Functionality "Diagnostic Panel" Mode with Logic Gates

Enabling Description: For a rapid, low-cost diagnostic application, the system is designed not for discovery but for answering a specific question (e.g., "Is this cell cancerous?"). A small panel of 5-10 detection reagents is used. The nucleic acid labels are not arbitrary barcodes but are engineered DNA strands that function as inputs to a DNA-based logic circuit (e.g., a strand displacement-based AND gate). For example, a cancer cell might be defined by the presence of Analyte A AND Analyte B BUT NOT Analyte C. The corresponding DNA labels (Label_A, Label_B, Label_C) are designed such that if Labels A and B are present in close proximity, they initiate a cascade that activates a fluorescent reporter. However, if Label_C is also present, it acts as an inhibitor, quenching the reporter. The final readout is not sequencing, but a simple fluorescence image where "ON" pixels indicate a positive diagnosis.

Mermaid Diagram:

graph TD
    subgraph Cell
        A[Analyte A Detected] --> L_A[Release Label_A]
        B[Analyte B Detected] --> L_B[Release Label_B]
        C[Analyte C Detected] --> L_C[Release Label_C]
    end
    subgraph LogicGate [DNA Logic Gate]
        L_A -- AND --> Z{Reporter Activation}
        L_B -- AND --> Z
        L_C -- NOT --> Z
    end
    Z --> F{Fluorescence ON/OFF};

Combination Prior Art Scenarios with Open-Source Standards

C.1. Integration with Micro-Manager and Open-Source Fluidics for Automated In-Situ Sequencing

Enabling Description: A complete, automated instrument for performing the reference technology's method is constructed using open-source components. The system utilizes an existing inverted microscope controlled by Micro-Manager (an open-source microscopy software package). A custom-built fluidics module, controlled by an Arduino MEGA board running open-source firmware, is integrated with the microscope's stage. The fluidics module consists of solenoid valves and a peristaltic pump to automatically handle the cyclical delivery of hybridization buffers, wash buffers, and sequencing reagents (polymerase, nucleotides, cleavage chemicals). A Python script within Micro-Manager synchronizes the fluidics (e.g., "flow nucleotide A") with image acquisition ("snap image"), automating the entire sequencing-by-synthesis process on the fixed sample. This disclosure provides the blueprint for a low-cost, reproducible hardware and software platform for spatial transcriptomics and proteomics.

C.2. A Standardized "Spatial-FastQ" (S-FQ) File Format for Spatially-Resolved Sequencing Data

Enabling Description: To standardize the data output from spatial sequencing technologies, a new file format, "Spatial-FastQ" or "S-FQ," is proposed as an extension of the ubiquitous FastQ format. A standard FastQ file contains a 4-line entry per read: Header, Sequence, Separator, and Quality. The S-FQ format modifies the header line to include mandatory fields for spatial coordinates. The proposed standard is: @Instrument:RunID:Flowcell:Lane:Tile:X:Y:Z ReadNum UMI=sequence. The X, Y, and Z fields store the micron-scale physical coordinates of the sequenced barcode within the tissue. This simple modification allows existing FastQ parsers to be easily adapted while enabling any downstream analysis tool to immediately link a sequence read to its point of origin without consulting separate lookup files, thereby standardizing and simplifying spatial omics data analysis.

C.3. ONNX-Compatible Machine Learning Models for Spatial Analyte Pattern Recognition

Enabling Description: A set of pre-trained machine learning models for interpreting the data generated by the reference technology is disclosed and made available in the ONNX (Open Neural Network Exchange) format. These models are trained on large, publicly available spatial atlases (e.g., the Human Cell Atlas). Disclosed models include:
1. A convolutional neural network (CNN) for identifying cell boundaries and performing cell segmentation based on a nuclear stain channel and total analyte signal.
2. A multi-layer perceptron (MLP) for classifying cell types (e.g., T-cell, macrophage, tumor cell) based on the vector of analyte counts within each segmented cell.
3. A graph attention network (GAT) for identifying significant ligand-receptor interactions between neighboring cells.
  By providing these models in ONNX format, they can be deployed across various deep learning frameworks (PyTorch, TensorFlow, etc.), democratizing access to advanced spatial data analysis.