Configurable handheld biological analyzers for identification of biological products based on raman spectroscopy

Abstract

To provide handheld biological analyzers and related biological analytics methods based on Raman spectroscopy.SOLUTION: A biological classification model configuration is loaded into a computer memory of a handheld biological analyzer. The biological classification model configuration includes a biological classification model configured to receive a Raman-based spectra dataset defining a biological product sample as scanned by a scanner. A spectral preprocessing algorithm is executed to reduce a spectral variance of the Raman-based spectra dataset. The biological classification model identifies a biological product type based on the Raman-based spectra dataset and further based on a classification component selected to reduce at least one of (1) a Q-residual error or (2) a summary-of-fit value of the biological classification model. The biological classification model configuration is transferrable to and loadable on other configurable handheld biological analyzers.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 62 / 925,893, filed October 25, 2019, and U.S. Provisional Patent Application No. 63 / 043,976, filed June 25, 2020. Each of the foregoing provisional patent applications is incorporated herein by reference in its entirety.

[0002] The present disclosure relates generally to configurable handheld biological analyzers, and more particularly to systems and methods for using configurable handheld biological analyzers for identifying or classifying biological products based on Raman spectroscopy. [Background technology]

[0003] The development and manufacture of pharmaceutical and biotechnology products commonly requires the measurement or identification of raw materials used in the development of such products. The purpose of product identity testing is to ensure the identity of the product. Situations where identity testing is necessary include product distribution to clinical sites, import testing, and transfer between network locations. Furthermore, the measurement or identification of biological products can be important to ensure the quality of the development or manufacturing process, and ultimately the quality of the final product itself, in order to meet quality standards and / or regulatory requirements.

[0004] The use of Raman spectroscopy for the measurement and identification of biological products is a relatively new concept. Generally, Raman spectroscopy can be used to probe the chemical or biological structure of raw materials or products. Raman spectroscopy is a non-destructive chemical or biological analytical technique that measures the interaction of light with a product or material, such as the biological attributes or chemical bonds of the product or material. Raman spectroscopy is a light scattering technique in which molecules in the sample substance or product scatter incident light from a high-intensity laser light source. Generally, most of the scattered light is of the same wavelength (color) as the laser source and does not provide any useful information. This is called Rayleigh scattering. However, a small amount of light is scattered at different wavelengths (colors) that are caused by the chemical or molecular structure of the material or product being analyzed. This is called Raman scattering and can be analyzed or scanned to generate Raman-based data of the material or product being analyzed.

[0005] Analysis of Raman scattering can provide detailed information about the properties of a material or product, including its chemical structure and / or identity, contamination and impurities, phase and polymorphism, crystallinity, intrinsic stress / strain, and / or intermolecular interactions. Such detailed information can be present in the Raman spectrum of the material. A Raman spectrum can be visualized to show many peaks across various wavelengths of light. A Raman spectrum can show the intensity and wavelength position of the Raman scattered light. Each peak can correspond to a specific molecular bond vibration associated with the material or product being analyzed.

[0006] In general, Raman spectra provide distinct chemical or biological "fingerprints" for particular materials, molecules, or products and can be used to verify the identity of a particular material, molecule, or product and / or distinguish it from others. Furthermore, Raman spectral libraries are often used to identify materials based on their Raman spectra. That is, Raman spectral libraries containing thousands of spectra can be searched to find a match with the Raman spectrum of a given material or product of interest, thereby identifying the given material or product.

[0007] Analytical devices currently exist that implement Raman spectroscopy to identify raw materials and products. For example, Thermo Fisher Scientific Inc. offers a handheld Raman-based analyzer that can be identified as the TruScan™ RM Handheld Raman Analyzer. However, the use of these existing scanners can be problematic due to discrepancies in scanning materials and / or products, such as pharmaceutical and biotechnology materials or products, especially those with similar Raman spectra. For example, due to discrepancies between the Raman spectra of similar products, existing handheld Raman-based analyzers can misidentify pharmaceutical or biotechnology products by outputting a Type I error (false positive) or a Type II error (false negative). The primary causes of discrepancies or errors stem from differences between Raman-based analyzers, including variations in software, manufacturing, age, components, operating environment (e.g., temperature), or other such differences between Raman-based analyzers.

[0008] Known approaches generally fail to address errors caused by inconsistencies or variations between handheld analyzers. For example, in one known approach, data from several analyzers can be used to develop static mathematical equations for use across several analyzers. However, a problem with this approach is that instrument performance may change over time. Also, it is often impractical or impossible to have routines that access all of these instruments. In particular, the data for building static mathematical equations is generally not available, especially for new analyzers, because manufacturers cannot provide new specifications for new analyzers in advance. This hinders the development and maintenance of static mathematical equations, especially as these new analyzers are developed over time and given that developing static mathematical equations typically requires a large number of samples for accuracy across various analyzer types. Furthermore, without these new specifications for new analyzers, static mathematical equations cannot be compatible when run on new analyzers. Furthermore, for example, differences in manufacturing or quality control of analytical equipment, particularly between various manufacturers, can result in excessive tolerance for variability in static mathematical equations, which in turn can produce static mathematical equations with excessive variability relative to accurate measurement and / or identification of biological products.

[0009] In a second known approach, data from a given analytical device is standardized and a child-parent device map is generated for a given group of analytical devices. However, this approach is limited because building a child-parent device map typically requires data from both the parent and child devices, which is typically difficult to implement and maintain and / or computationally expensive, especially over time as new generations of analytical devices are developed, thereby requiring numerous permutations and variations of child-parent device maps. Furthermore, in the biopharmaceutical industry, user access to child devices is limited, which also limits the child-parent device map approach.

[0010] A third known approach also standardizes data from a given analyzer, but in this case analyzer-to-analyzer variability is ignored or treated as a minor issue, which is undesirable given that analyzer-to-analyzer variability typically affects the accurate identification and measurement of raw materials and / or biological products and therefore should be taken into account.

[0011] For the foregoing reasons, a need exists for a configurable handheld biological analyzer and related methods for identifying biological products based on Raman spectroscopy that are configured to reduce variability and increase interchangeability between similarly configured configurable handheld biological analyzers. Summary of the Invention [Means for solving the problem]

[0012] The present disclosure describes the use of Raman spectroscopy via handheld analytical devices to identify biological products. Furthermore, the disclosure herein describes the use of configurable handheld biological analytical devices, systems, and methods that overcome limitations generally associated with known uses of Raman spectroscopy to measure biological products. For example, certain biological products may have Raman spectra that are too similar to be distinguishable by known methods using Raman spectroscopy, which generally rely on general-purpose statistical algorithms. Raman spectral measurements can be particularly problematic when instrument-to-instrument variability is introduced, resulting in, for example, Type I and Type II errors between various analytical devices. As described herein, such variability can arise from any one or more of differences in software, manufacturing, age, components, operating environment (e.g., temperature), or other differences between Raman-based analytical devices. This issue is particularly evident during the development or manufacturing of biological products, as instrument-to-instrument variability can be a key factor affecting the quality, robustness, and / or transferability of processes related to pharmaceutical or biological products. Thus, in various embodiments disclosed herein, configurable handheld biological analytical devices are described that employ configurations that employ specific pre-processing algorithms and / or multivariate data analysis to, for example, (1) ensure that measurement and / or identification of materials or products is sensitive and / or specific, and (2) ensure that, when deployed on a first set of analytical devices, the compatibility and configuration is transferable and / or implementable to additional analytical devices, such as new analytical devices within a "network" or group of analytical devices.

[0013] Thus, various embodiments herein disclose a configurable handheld biological analyzer for identifying biological products based on Raman spectroscopy. The configurable handheld biological analyzer may include a first housing adapted for handheld operation. Furthermore, the configurable handheld biological analyzer may include a first scanner carried by the first housing. The configurable handheld biological analyzer may include a first processor communicatively coupled to the first scanner. The configurable handheld biological analyzer may further include a first computer memory communicatively coupled to the first processor. In various embodiments, the first computer memory may be configured to load a biological classification model configuration. The biological classification model configuration may include a biological classification model. The biological classification model may be configured to execute on the first processor. The first processor may be configured to (1) receive a first Raman-based spectral dataset defining a first biological product sample scanned by the first scanner, and (2) identify a biological product type based on the first Raman-based spectral dataset using the biological classification model. The biological classification model configuration may include a spectral pre-processing algorithm. The first processor may be configured to execute the spectral pre-processing algorithm when the first Raman-based spectral dataset is received by the first processor to reduce spectral discrepancies in the first Raman-based spectral dataset. Further, the biological classification model may include a classification component selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model may be configured to identify a biological product type of the first biological product sample based on the classification component.

[0014] In an additional embodiment disclosed herein, a bioanalytical method for identifying a biological product based on Raman spectroscopy is disclosed. The bioanalytical method may include loading a biological classification model configuration into a first computer memory of a first configurable handheld biological analytical device having a first processor and a first scanner. The biological classification model configuration may include a biological classification model. The bioanalytical method may further include receiving a first Raman-based spectral dataset that defines a first biological product sample scanned by the first scanner according to the biological classification model. The bioanalytical method may further include executing a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the first Raman-based spectral dataset. Still further, the bioanalytical method may include identifying a biological product type based on the first Raman-based spectral dataset using the biological classification model. The biological classification model may include classification components selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of the fit of the biological classification model, and the biological classification model may be configured to identify a biological product type of the first biological product sample based on the classification components.

[0015] In yet a further embodiment disclosed herein, a tangible, non-transitory, computer-readable medium (e.g., computer memory) storing instructions for identifying biological products based on Raman spectroscopy is described. The instructions, when executed by one or more processors of a configurable handheld biological analyzer, cause the one or more processors of the configurable handheld biological analyzer to load a biological classification model configuration into the computer memory of the configurable handheld biological analyzer having a scanner. The biological classification model configuration may include a biological classification model. The biological classification model can receive a dataset of Raman-based spectra defining a biological product sample scanned by the scanner. Further, the one or more processors of the configurable handheld biological analyzer can execute a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the dataset of Raman-based spectra. The one or more processors of the configurable handheld biological analyzer can use the biological classification model to identify a biological product type based on the dataset of Raman-based spectra. As described in various embodiments, the biological classification model may include classification components selected to reduce at least one of: (1) a Q residual error of the biological classification model; or (2) a summary value of the fit of the biological classification model. The biological classification model may be configured to identify a biological product type of a biological product sample based on the classification components.

[0016] Advantages of the present application include the development of biological classification models (e.g., multivariate analytical models) that provide consistent results for the same pharmaceutical or biological product (e.g., therapeutic product / agent) across a variety of analytical devices, including a variety of analytical devices used to scan the Raman-based datasets used to build the biological classification model. As described herein, multiple analytical devices, or multiple datasets of Raman spectra generated by such analytical devices, can be used to build the biological classification model.

[0017] Furthermore, as described herein, the biological classification models are configurable and transferable between configurable handheld biological analyzers, which may include Raman spectral preprocessing, classification component selection (e.g., via singular value decomposition (SVD) analysis), and discriminative statistical analysis to reduce variability between configurable handheld biological analyzers. For example, use of the biological classification models described herein improves existing analyzers by reducing instrument / analyzer variability, not requiring instrument data for development, and being usable across different analyzers implementing different software, having different software or software versions, having different makes, ages, operating environments (e.g., temperature), components, or other such differences.

[0018] In various embodiments, the Q-residual can be used as a discriminative statistic to determine which biological classification models tolerate analyzer-to-analyzer variability, which can provide guidance on which biological classification models to select for loading onto a configurable handheld biological analyzer.

[0019] Furthermore, the accuracy of the biological classification model may be increased by applying preprocessing techniques (e.g., the spectral preprocessing algorithms described herein) to minimize statistical Type I and / or Type II errors in the output of the biological classification model and therefore improve the output of the configurable handheld biological analyzer in which the biological classification model is implemented / configured.

[0020] Additionally, in some embodiments, the configurable handheld biological analyzer can use multivariate analysis (e.g., principal component analysis (PCA)) to determine classification components of the biological classification model. This allows the configurable handheld biological analyzer to distinguish between biological products / agents with similar formulations. This provides a flexible approach, as biological classification models can be generated that include various, different, and / or additional classification components (e.g., a second principal component in a PCA biological classification model) that correspond to products with multiple specifications (e.g., products for denosumab).

[0021] In accordance with the above and the disclosure herein, the present disclosure includes improvements to the functionality of a computer or improvements to other technologies, at least as the present claims recite, for example, a configurable handheld biological analyzer for identifying biological products based on Raman spectroscopy, which is an improvement over existing handheld biological analyzers. That is, because the configurable handheld biological analyzer, as described herein, is a computing device, and through its biological classification model configuration, provides reduced analyzer-to-analyzer variability compared to existing handheld biological analyzers, the present disclosure describes improvements to the functionality of the computer itself or "any other technology or technical field." This means that, at a minimum, the configurable handheld biological analyzer described herein improves upon the prior art, providing increased accuracy with respect to the measurement, identification, and / or classification of materials and / or products (e.g., therapeutic products), a key feature in the manufacturing and development of pharmaceutical and / or other such biological products.

[0022] Additionally, the configurable handheld biological analyzers described herein are further improved by using biological classification model configurations that can be transferred to, optionally updated (with new data), and loaded into the memory of compatible configurable handheld biological analyzers, allowing for standardization among a set (i.e., "network") of analyzers, thereby reducing variability. This reduces maintenance and / or deployment time of configurable handheld biological analyzers for an analyzer network.

[0023] Additionally, the configurable handheld biological analyzer is further improved by using a biological classification model configuration including a biological classification model, as described herein, that improves the accuracy of identification and / or classification of biological products by eliminating or reducing Type I errors (e.g., false positives) and / or Type II errors (e.g., false negatives).

[0024] Additionally, the present disclosure includes applying certain claim elements with or by using certain machines, such as configurable handheld biological analyzers, to identify biological products based on Raman spectroscopy, including identifying biological products during the development or manufacturing of such products.

[0025] Additionally, the present disclosure includes transforming or reducing a particular item to a different state or object, for example, transforming or reducing a Raman spectral data set to a different state that is used to identify biological products based on Raman spectroscopy.

[0026] The present disclosure includes specific features other than routine conventional activities well understood in the art, or the addition of non-conventional steps that limit the scope of the present claims to particular beneficial applications (e.g., including providing biological classification model configurations used to reduce variability among a set (i.e., a "network") of configurable handheld biological analytical devices, each of which may be used to identify biological products based on Raman spectroscopy).

[0027] Advantages will become more apparent to those skilled in the art from the following description of preferred embodiments, which are shown and described by way of illustration. As will be understood, the present embodiments are capable of other and different embodiments, and their details are capable of modification in various respects. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive.

[0028] The drawings described below depict various aspects of the systems and methods disclosed therein. It should be understood that each drawing depicts one embodiment of a particular aspect of the disclosed systems and methods, and that each drawing is intended to correspond to a possible embodiment thereof. Furthermore, wherever possible, the following description refers to reference numerals included in the drawings described below, and features depicted in multiple figures are designated with consistent reference numerals.

[0029] The drawings show arrangements that are discussed herein, it being understood, however, that the present embodiments are not limited to the precise arrangements and instrumentalities shown. [Brief explanation of the drawings]

[0030] [Figure 1] 1 illustrates an example of a configurable handheld biological analyzer for identifying biological products based on Raman spectroscopy, according to various embodiments disclosed herein. [Figure 2] 1 shows an exemplary flow chart of a bioanalytical method for identifying biological products based on Raman spectroscopy, according to various embodiments disclosed herein. [Figure 3A] 1A-1C illustrate exemplary visualizations of Raman-based spectral datasets scanned by various handheld biological analyzers according to various embodiments disclosed herein. [Figure 3B] 3B shows an exemplary visualization of a modified Raman-based spectral dataset modified from the Raman-based spectral dataset of FIG. 3A. [Figure 3C] 3C shows an exemplary visualization of a normalized Raman-based spectral dataset as a normalized version of the modified Raman-based spectral dataset of FIG. 3B. [Figure 4A] 1 shows an exemplary visualization of the Q-residual error of a biological classification model. [Figure 4B] 10 shows an exemplary visualization of a summary value of the fit (e.g., Hotelling T^2 value) of a biological classification model. [Figure 5] 1 shows an exemplary visualization of Raman spectra of biological product types according to various embodiments disclosed herein. [Figure 6A] 1 shows an exemplary computer program listing containing pseudocode for biological classification model construction according to various embodiments disclosed herein. [Figure 6B] 1 shows an exemplary computer program listing containing pseudocode for biological classification model construction according to various embodiments disclosed herein. [Figure 6C] 1 shows an exemplary computer program listing containing pseudocode for biological classification model construction according to various embodiments disclosed herein. [Figure 7] 10 illustrates an example visualization of reduced Q residual error according to various embodiments described herein. [Figure 8A] 10 shows an exemplary visualization of reduced Q residual error for target products evaluated for each of 18 different configurable handheld biological analyzers according to various embodiments described herein. [Figure 8B]10 shows an exemplary visualization of reduced Q residual error for target products evaluated for each of 18 different configurable handheld biological analyzers according to various embodiments described herein. [Figure 8C] 10 shows an exemplary visualization of reduced Q residual error for target products evaluated for each of 18 different configurable handheld biological analyzers according to various embodiments described herein. [Figure 8D] 10 shows an exemplary visualization of reduced Q residual error for target products evaluated for each of 18 different configurable handheld biological analyzers according to various embodiments described herein. [Figure 8E] 1 shows an exemplary visualization of target product reduced fit summary values evaluated for 18 different configurable handheld biological analyzers according to various embodiments described herein. DETAILED DESCRIPTION OF THE INVENTION

[0031] The drawings depict preferred embodiments for purposes of illustration only. Alternative embodiments of the systems and methods illustrated herein may be employed without departing from the principles of the invention described herein.

[0032] FIG. 1 illustrates an example of a configurable handheld biological analyzer 102 for identifying a biological product 140 based on Raman spectroscopy, according to various embodiments disclosed herein. In the embodiment of FIG. 1, the configurable handheld biological analyzer 102 includes a first housing 101 shaped or adapted for handheld operation. Additionally, the configurable handheld biological analyzer 102 includes a first scanner 106 carried by (e.g., directly or indirectly coupled or connected to) the first housing. The configurable handheld biological analyzer 102 also includes a first processor 110 communicatively coupled to the first scanner 106. The configurable handheld biological analyzer 102 may further include a first computer memory 108 communicatively coupled to the first processor 110. Additionally, the configurable handheld biological analyzer 102 may include an input / output (I / O) component 109 for receiving input from a navigation wheel 105. For example, a user can manipulate navigation wheel 105 to select or scroll through data or information for a particular sample of a scanned biological product, e.g., from scanning biological product 140. Input / output (I / O) component 109 can also control the display of measurement, identification, classification, or other information described herein on display screen 104. Display screen 104, navigation wheel 105, first scanner 106, first computer memory 108, I / O component 109, and / or first processor 110 are each communicatively coupled over electronic bus 107 configured to send and / or receive electronic signals (e.g., control signals) or information between the various components, including 104-110. In some embodiments, configurable handheld biological analyzer 102 can be a Raman-based handheld analyzer, such as a TruScan™ RM Handheld Raman Analyzer offered by Thermo Fisher Scientific Inc.

[0033] In various embodiments, the first computer memory 108 is configured to load a biological classification model configuration, such as biological classification model configuration 103. The biological classification model configuration 103 may be used to implement the biological analysis method of FIG. 2 for identifying biological products based on Raman spectroscopy, as further described herein.

[0034] In the embodiment of FIG. 1, the biological classification model configuration 103 is implemented as an XML file in Extensible Markup Language (XML) format. As described in various embodiments herein, FIGS. 6A-6C show exemplary XML-formatted computer program listings including pseudocode for a biological classification model configuration (e.g., biological classification model configuration 103). In the computer program listings of the embodiment of FIGS. 6A-6C, for example, in code section 1, the biological classification model configuration 103 is formatted in XML, where the biological classification model (" <model>" ) is defined within biological classification model configuration 103. Biological classification model configuration 103 is transferable, installable, and / or implementable or executable on similarly configured configurable handheld biological analyzer devices (e.g., configurable handheld biological analyzer devices 112, 114, and / or 116). Configurable handheld biological analyzer devices 112, 114, and 116 each include the same components as configurable handheld biological analyzer device 102, and therefore the disclosure of configurable handheld biological analyzer device 102 applies equally to each of configurable handheld biological analyzer devices 112, 114, and 116. Configurable handheld biological analyzer devices 102, 112, 114, and 116 may each be part of the same group or set of analyzers (i.e., including a "network" or group of analyzers). In some embodiments, the configurable handheld biological analyzers 102, 112, 114, and / or 116 may each have a mix of the same, similar, and / or different properties or characteristics, such as the same, similar, and / or different software versions or types, manufactures, ages, operating environments (e.g., temperatures), components, or other similarities or differences of such Raman-based analyzers.

[0035] Regardless of the mix of the same, similar, and / or different characteristics or features among the configurable handheld biological analytical devices 102, 112, 114, and 116, the biological classification model configuration 103 and its associated biological classification model enable a network of configurable handheld biological analytical devices (e.g., configurable handheld biological analytical devices 102, 112, 114, and 116) to produce consistent results when measuring or identifying pharmaceutical or biological products (e.g., therapeutic products / agents). That is, regardless of the similarities or differences among a given analytical device network of configurable handheld biological analytical devices, when such configurable handheld biological analytical devices are configured with the biological classification model configurations described herein, such configurable handheld biological analytical devices can accurately identify or measure a given pharmaceutical or biological product.

[0036] In various embodiments, multiple analytical devices may be used to generate or build the biological classification model configuration 103 and its associated biological classification models. For example, in some embodiments, any one or more of the configurable handheld biological analytical devices 102, 112, 114, and 116 and / or other analytical devices (not shown) may be used to generate or build the biological classification models.

[0037] Generating the biological classification model configuration 103 and its associated biological classification model generally requires a group or network of analytical devices to scan samples (e.g., of biological products 140) to create a dataset of Raman-based spectra for those samples. For example, scanning the biological product 140, e.g., by any of the configurable handheld biological analytical devices 102, 112, 114, and 116, can provide detailed information about the biological product 140. For example, the detailed information can include a dataset of Raman-based spectra defining the biological product sample (e.g., of the biological product 140). Examples of biological products 140 can include any of the denosumab DP, panitumumab DP, etanercept DP, pegfilgrastim DP, romosozumab DP, adalimumab DS, and / or erenumab DP described herein (e.g., romosozumab DP, adalimumab DS, and / or erenumab DP). However, it should be understood that additional biological products are contemplated herein and that biological product 140 is not limited to any particular biological product or category thereof.

[0038] In some embodiments, the configurable handheld biological analyzer 102 can define instrument- or analyzer-based spectral acquisition parameters (e.g., integration time, laser power, etc.) for use in scanning a sample of, for example, biological product 140. For example, a user can select the spectral acquisition parameters for use in scanning a sample via the navigation wheel 105. In some embodiments, the configurable handheld biological analyzer 102 can generate an output file (e.g., an output file in the ".acq" file format) that specifies the spectral acquisition parameters.

[0039] In some embodiments, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) can load an output file (e.g., an ".acq" file) to configure the configurable handheld biological analyzer with spectral acquisition parameters for use in scanning a target product. As described herein, a dataset of Raman-based spectra can be scanned by one or more configurable handheld biological analyzers (e.g., configurable handheld biological analyzer 102) to generate a biological classification model configuration (e.g., biological classification model configuration 103). In some embodiments, samples (e.g., multiple lots) of a biological product (e.g., biological product 140) can be selected as representative target products for scanning. Generally, as described herein, "target product" refers to a biological product used to train or configure a biological classification model configuration and its associated model. Generally, a target product is selected based on its biological specifications. Once set up with spectral acquisition parameters for use in scanning the target product, the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) can scan the target product sample (e.g., with the first scanner 106), possibly multiple times (e.g., 14 times), where each scan generates detailed information including a dataset of the Raman-based spectrum of the target product.

[0040] In a similar embodiment, multiple configurable handheld biological analyzers (configurable handheld biological analyzers 102, 112, 114, and / or 116) can be set up with spectral acquisition parameters to load an output file (e.g., an ".acq" file) and use each configurable handheld biological analyzer to scan a biological product sample. Once set up, each configurable handheld biological analyzer (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116) is configured to scan the sample (e.g., with first scanner 106), possibly multiple times (e.g., 14 times), where each scan generates detailed information including a dataset of Raman-based spectra of the target product. By scanning a given target product with different / multiple scanners, the Raman-based spectral datasets acquired by these scanners are robust in that they capture any variations between scanners (e.g., arising from software, manufacturing, age, operating environment (e.g., temperature), etc.). In this manner, the Raman-based spectral datasets provide an ideal training dataset for reducing the variability between multiple scanners described herein. For example, each of the Raman-based spectral datasets scanned by multiple scanners (e.g., any of the configurable handheld biological analyzers 102, 112, 114, and / or 116) may be output and / or saved as a Raman spectral file, e.g., having an ".spc" file format.

[0041] It should be understood that a Raman-based spectral dataset may also be obtained for a challenge product in the same or similar manner as for a target product. As used herein, "challenge product" refers to a biological product (e.g., selected from biological products 140) that a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) is configured to identify, classify, or measure when loaded with or configured with a biological classification model configuration (e.g., biological classification model configuration 103) and its associated biological classification model, as described herein.

[0042] The Raman-based spectral dataset for the challenge product can be acquired in the same or similar manner as the target product, where the challenge product can be selected based on its biological specification, and the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) can be configured with spectral acquisition parameters to load an output file (e.g., an ".acq" file) and use the configurable handheld biological analyzer to scan the challenge product. Once set up, the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) is configured to scan the challenge product sample (e.g., with the first scanner 106), possibly multiple times (e.g., three times), where each scan generates detailed information including the Raman-based spectral dataset for the challenge product. For example, the Raman-based spectral dataset scanned by the configurable handheld biological analyzer 102 can be output and / or saved as a Raman spectral file, e.g., having an ".spc" file format.

[0043] In some embodiments, generation of the biological classification model configuration (e.g., biological classification model configuration 103) may be performed by a remote processor, such as the processor of computer 130 shown in FIG. 1 . For example, a Raman-based spectral dataset generated for a biological product described herein (e.g., selected from biological products 140) may be imported and / or analyzed by modeling software running on computer 130 configured to analyze the Raman-based spectral dataset. One example of such modeling software is SOLO (stand-alone chemometrics software) provided by Eigenvector Research, Inc. However, it should be understood that other modeling software, including custom or proprietary software implemented to perform the functions described herein, may also be used. The modeling software may construct or generate the biological classification model based on the Raman-based spectral dataset. For example, in some embodiments, a Raman-based spectral dataset scanned or acquired for a target product described herein may be used to construct or generate the biological classification model. Furthermore, a Raman-based spectral dataset (e.g., of a target product or a challenge product) may be used for cross-validation of the biological classification model. For example, a Raman-based spectral dataset can be used to assess Type I errors (e.g., false positives) and Type II errors (e.g., false negatives) of a biological classification model against a Raman-based spectral dataset cross-validation dataset.

[0044] In various embodiments, a biological classification model and / or its associated biological classification model configuration (e.g., biological classification model configuration 103) may be generated that includes algorithms (e.g., scripts) and parameters used by a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) to identify, classify, and / or measure biological products as described herein. Examples of algorithms (e.g., scripts) and / or parameters are described herein in connection with FIGS. 2, 6A, and 6B. For example, a biological classification model configuration (e.g., biological classification model configuration 103) may include parameters that define the details of the biological classification model. For example, such parameters may determine the number of classification components, loadings, etc. of the biological classification model. For example, in one embodiment, the number of classification components may be determined by the modeling software via, e.g., singular value decomposition (SVD) analysis, where the classification components include one or more principal components of PCA. The modeling software may be configured to set a statistical confidence level for determining the classification components (e.g., principal components) to include in the biological classification model. For example, in the computer program listing embodiment of FIGS. 6A-6C, in code section 1, the biological classification model configuration is <model>") indicates a PCA-type biological classification model. This indicates that the classification components of the biological classification model will be principal components. For example, in the embodiment of FIGS. 6A-6C, code section 2 indicates that the number of principal components will be one (single) principal component ("Num.PCs:1"), determined, for example, via an SVD analysis ("Algorithm:SVD") performed on the first processor 110 of the configurable handheld biological analyzer 102.

[0045] As a further example, a biological classification model configuration (e.g., biological classification model configuration 103) may include computer code or script for defining or implementing a spectral pre-processing algorithm, e.g., as described in connection with Figures 3A-3C. Generally, the computer code or script for defining or implementing the spectral pre-processing algorithm may be executed on a processor (e.g., first processor 110), where the processor receives a dataset of Raman-based spectra of a biological product (e.g., biological product 140). The configurable handheld biological analyzer subsequently executes the computer code or script for defining or implementing the spectral pre-processing algorithm to prepare / pre-process the data for input into the classification component of the biological classification model in order to identify, measure, or classify a biological product (e.g., a challenge product) as described herein. For example, in the computer program listing embodiment of Figures 6A-6C, in code section 2, the biological classification model configuration includes an execution order (e.g., "Preprocessing: 1st Derivative (order: 2, window: 21 pt, incl only, tails: polyinterp), SNV, Mean Center") of an exemplary spectral pre-processing algorithm that includes determining a first derivative, applying a standard normal variate (SNV) algorithm, and further applying a mean centering function to a dataset of scanned Raman-based spectra for a particular product (e.g., a target product or a challenge product). An exemplary embodiment of this execution order is described and visualized herein in connection with Figures 3A-3C and code sections 4-6 of Figures 6A-6C.

[0046] As a further example, a biological classification model configuration (e.g., biological classification model configuration 103) may include a dataset of Raman-based spectra used to generate the biological classification model. For example, in the computer program listing embodiment of Figures 6A-6C, in code section 3, a biological classification model configuration (e.g., biological classification model configuration 103) includes an exemplary dataset of Raman-based spectra used to generate the biological classification model of Figures 6A-6C.

[0047] In some embodiments, the biological classification model configuration (e.g., biological classification model configuration 103) may also define thresholds as statistical acceptance criteria, for example, for determining whether a biological product has been successfully identified or measured by the configurable handheld biological analytical device 102. For example, such thresholds may be used to determine whether a biological product has been successfully identified or measured by the configurable handheld biological analytical device 102, such as the Q residual or the Hotelling T 2 Pass / fail threshold values (as described herein) can be defined. In other embodiments, the thresholds can be configured independently of the biological classification model configuration (e.g., biological classification model configuration 103), for example, by a user manually configuring and / or defining the thresholds via the navigation wheel 105 and display screen 104 described herein.

[0048] Once generated, the biological classification model and its associated biological classification model configuration (e.g., biological classification model configuration 103) may be exported to a file (e.g., an XML file as described herein) for transmission (e.g., via computer network 120 or in other manners as described herein) to a configurable handheld biological analysis device (e.g., any one or more of configurable handheld biological analysis devices 102, 112, 114, and / or 116) as described herein and / or for loading into memory of a configurable handheld biological analysis device (e.g., any one or more of configurable handheld biological analysis devices 102, 112, 114, and / or 116) as described herein. In some embodiments, the output file (e.g., an ".acq" file as described herein) may also be transmitted (e.g., via computer network 120 or in other manners described herein) to a configurable handheld biological analyzer (e.g., any one or more of configurable handheld biological analyzers 102, 112, 114, and / or 116) and / or loaded into memory of the configurable handheld biological analyzer (e.g., any one or more of configurable handheld biological analyzers 102, 112, 114, and / or 116).

[0049] The biological classification models may be generated by a remote processor that is remote from a given configurable handheld biological analyzer device. For example, in the embodiment of FIG. 1 , computer 130 includes a remote processor that is remote from configurable handheld biological analyzer device 102. Computer 130 can generate (e.g., as described herein) one or more biological classification model configurations and / or biological classification models to store in database 132. In various embodiments, computer 130 can transfer a biological classification model configuration (e.g., any of biological classification model configurations 103, 113, 115, and / or 117) to a configurable handheld biological analyzer device (e.g., to configurable handheld biological analyzer devices 102, 112, 114, and / or 116, respectively) over computer network 120. In some embodiments, each of biological classification model configurations 103, 113, 115, and / or 117 can be a copy of the same file (e.g., the same XML file). The computer network 120 may include wired and / or wireless (e.g., 802.11 standard networks) implementing computer packet protocols such as Transmission Control Protocol (TCP) / Internet Protocol (IP). In other embodiments, the biological classification model configuration (e.g., the biological classification model configuration 103) may be transferred via a Universal Serial Bus (USB) cable (not shown), a memory drive (e.g., flash or thumb drive) (not shown), a disk (not shown), or other transfer or memory device capable of transferring data files, such as the XML files disclosed herein. In still further embodiments, the biological classification model configuration 103 may be transferred via wireless standards or protocols, such as Bluetooth, WiFi, or cellular standards such as GSM, EDGE, CDMA, etc.

[0050] Biological classification model configurations (e.g., biological classification model configuration 103) can be transferred between configurable handheld biological analyzers. Once transferred, the biological classification model configuration can be loaded into the memory of the configurable handheld biological analyzer to calibrate or configure the handheld biological analyzer to reduce variability relative to other configurable handheld biological analyzers that implement or execute the biological classification model. For example, in one embodiment, the biological classification model configuration 103 can include a biological classification model. The biological classification model of the biological classification model configuration 103 can be configured to execute on the first processor 110. For example, the first processor 110 can be configured to (1) receive a first Raman-based spectral dataset defining a first biological product sample scanned by a first scanner (e.g., of a scan of a biological product 140), and (2) use the biological classification model to identify a biological product type based on the first Raman-based spectral dataset. For example, in some embodiments, the biological product type can be that of a therapeutic product having a therapeutic product type.

[0051] The biological classification model of biological classification model configuration 103 may be electronically transferred, for example, over computer network 120 via biological classification model configuration 113 to configurable handheld biological analyzer 112. With respect to configurable handheld biological analyzer 102 only, configurable handheld biological analyzer 112 may include a second housing adapted for handheld operation, a second scanner coupled to the second housing, a second processor communicatively coupled to the second scanner, and a second computer memory communicatively coupled to the second processor. The second computer memory of configurable handheld biological analyzer 112 is configured to load biological classification model configuration 113. Biological classification model configuration 113 includes the biological classification model of biological classification model configuration 103. When implemented or executed on the second processor of the configurable handheld biological analyzer 112, the second processor is configured to (1) receive a second Raman-based spectral dataset defining a second biological product sample scanned by a second scanner of the configurable handheld biological analyzer 112 (e.g., taken from scanning a biological product 140), and (2) use a biological classification model to identify a biological product type based on the second Raman-based spectral dataset. In such an embodiment, the same biological classification model transferred by the biological classification model configuration file is used to identify the same biological product or product type, where the second biological product sample is a new sample of this biological product type (e.g., the same biological product type as that analyzed by the first configurable handheld biological analyzer 102).

[0052] In various embodiments, new or additional Raman-based spectral datasets may be scanned by a configurable handheld biological analyzer and used to update the biological classification model. In such embodiments, the updated biological classification model may be transferred to a configurable handheld biological analyzer described herein (e.g., configurable handheld biological analyzer 102).

[0053] In some embodiments, a computer memory (e.g., first computer memory 108) of a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) may be configured to load a new biological classification model, where the new biological classification model may include updated classification components. The new classification components may be generated or determined, for example, for a new biological classification model received along with a new biological classification model configuration (e.g., biological classification model configuration 103).

[0054] As described in various embodiments herein, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) can be configured by loading a biological classification model configuration and its associated biological classification model. Once configured, the configurable handheld biological analyzer 102 can be used to identify, classify, or measure a product of interest (e.g., a challenge product and / or sample) as described herein.

[0055] FIG. 2 illustrates an exemplary flowchart of a biological analytical method 200 for identifying a biological product (e.g., biological product 140) based on Raman spectroscopy, according to various embodiments disclosed herein. The biological analytical method 200 begins (202) at block 204 by loading a biological classification model configuration (e.g., biological classification model configuration 103) into a first computer memory (e.g., first computer memory 108) of a first configurable handheld biological analytical device having a first processor (e.g., first processor 110) and a first scanner (e.g., first scanner 106). In the embodiment of FIG. 2, the biological classification model configuration (e.g., biological classification model configuration 103) includes a biological classification model described herein. Additionally, in some embodiments, the configurable handheld biological analytical device (e.g., configurable handheld biological analytical device 102) can load, for example, into memory 108, spectral acquisition parameters (e.g., in an ".acq" file) for use in scanning the product.

[0056] In block 206, the biological analysis method 200 includes receiving a first Raman-based spectral dataset that defines a first biological product sample (e.g., selected from biological products 140) scanned by a first scanner (e.g., first scanner 106) according to a biological classification model (e.g., of biological classification model configuration 103).

[0057] In block 208, the biological analysis method 200 includes, for example, executing, by a processor (e.g., first processor 110), a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the first Raman-based spectral dataset. Spectral discrepancies refer to inter-analytical device spectral discrepancies between the first Raman-based spectral dataset and one or more other corresponding Raman-based spectral datasets of one or more other handheld biological analyzers. For example, spectral discrepancies may exist between the Raman-based spectral dataset scanned by configurable handheld biological analyzer 102 and the Raman-based spectral dataset scanned by configurable handheld biological analyzer 112. Spectral discrepancies may exist even when the Raman-based spectral datasets scanned by each analytical device represent the same biological product type. Such spectral discrepancies may arise due to analytical device-to-analytical device variations and / or differences, such as different software versions, manufacturing, age, operating environment (e.g., temperature), components, or other differences among the Raman-based analytical devices described herein.

[0058] The spectral pre-processing algorithm is configured to reduce inter-analyzer spectral discrepancies between the first Raman-based spectral dataset and one or more other Raman-based spectral datasets. For example, in various embodiments, the spectral pre-processing algorithm is implemented or executed (e.g., on first processor 110) to minimize statistical Type I (e.g., false positives) and / or Type II (e.g., false negatives) errors associated with identifying a biological product (e.g., biological product 140). In various embodiments, the spectral pre-processing algorithm may reduce inter-analyzer spectral discrepancies among multiple configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116).

[0059] 3A-3C illustrate an exemplary execution sequence of a spectral pre-processing algorithm for a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102). Execution of the spectral pre-processing algorithm (e.g., by first processor 110) mitigates or mitigates the effects of variations inherent in each analyzer (e.g., configurable handheld biological analyzer 102, 112, 114, and / or 116) and reduces discrepancies between Raman-based spectral datasets generated by scans of these analyzers. FIG. 3A illustrates a visualization 302 of exemplary Raman-based spectral datasets (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) scanned by one or more handheld biological analyzers, according to various embodiments disclosed herein. The Raman-based spectral dataset of Figure 3A may include the Raman-based spectral datasets (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) used to generate the biological classification model configurations (e.g., biological classification model configuration 103) and their associated biological classification models described herein. For example, the Raman-based spectral dataset of Figure 3A may be the one identified in code section 3 of Figure 6A.

[0060] In some embodiments, each of the Raman-based spectral datasets in Figure 3A (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) may represent a scan of a different configurable handheld biological analyzer (e.g., any of configurable handheld biological analyzer devices 102, 112, 114, and / or 116). However, in other embodiments, each of the Raman-based spectral datasets in Figure 3A (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) may represent multiple scans of the same configurable handheld biological analyzer (e.g., configurable handheld biological analyzer device 102).

[0061] FIG. 3A shows several Raman-based spectral datasets (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) visualized across Raman intensity values (on a Raman intensity axis 304) and optical wavelength / frequency values (on a Raman shift axis 306). The Raman intensity axis 304 shows the intensity of scattered light at a given wavelength across the Raman shift axis 306. The Raman intensity axis 304 can show the number of photons scattered by a biological product sample when scanned by an analytical device (e.g., a configurable handheld biological analyzer 102) (e.g., a data / value of 3 is a relative indication of the intensity of the photons measured / scanned by the first scanner 106). The Raman shift axis 306 shows the wavenumber (e.g., inverse wavelength) of the scattered light. The units of wavenumber (i.e., number of waves per centimeter (cm), cm) are used. -1 ) indicates the difference in frequency or wavelength between the incident light and the scattered light. In the visualization 302 of FIG. 3A, the shift axis 306 is from 600 to 1500 cm -1 The Raman intensity axis 304 includes a Raman intensity range of 1 to 5. As shown in FIG. 3A, each of the Raman-based spectral datasets (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) includes a Raman intensity range of 600 to 1500 cm. -1 Visualize the measured Raman intensity values over the optical spectral range.

[0062] Furthermore, in various embodiments, each of the Raman-based spectral datasets of Figure 3A (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) may represent scans of the same biological product sample having the same biological product type. In such embodiments, as shown in Figure 3A, there will be variation in the Raman intensity values (on Raman intensity axis 304) of the Raman-based spectral datasets (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) across the optical wavelength / frequency values (on Raman shift axis 306), even though any one or more of the configurable handheld biological analyzers may have scanned the same biological product sample having the same biological product type. As described herein, variations may arise due to software, manufacturing, age, optical components, operating environment (e.g., temperature), or differences between configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116).

[0063] FIG. 3B shows an exemplary visualization 312 of a modified Raman-based spectral dataset modified from the Raman-based spectral dataset of FIG. 3A. For example, FIG. 3B may represent a first stage in the execution sequence of a spectral pre-processing algorithm. The visualization 312 of FIG. 3B includes the same Raman intensity axis 304 and Raman shift axis 306 described herein with respect to FIG. 3A. In the embodiment of FIG. 3B, a processor (e.g., first processor 110) applies a derivative transform to the Raman-based spectral dataset of FIG. 3A (e.g., including Raman-based spectral datasets 302a, 302b, and 302c) to generate the modified Raman-based spectral dataset (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) shown in FIG. 3B. Specifically, in the embodiment of FIG. 3B , a first derivative is applied that includes 11-15 point data smoothing (i.e., determining a Raman weighted average of a contiguous group of 11-15 Raman shift values, followed by applying a first derivative transform to this group). In other words, the derivative transform shown in FIG. 3B involves a processor (e.g., first processor 110) determining a Raman weighted average (of Raman intensity axis 304) of a contiguous group of 11-15 Raman shift values across Raman shift axis 306, followed by a processor (e.g., first processor 110) determining the corresponding derivative of these Raman weighted averages across Raman shift axis 306. Applying the derivative transform mitigates the effects of background curvature, for example, due to Rayleigh scattering / rejection optics and / or other dispersive elements. This is illustrated by a comparison of visualization 302 in FIG. 3A with visualization 312 in FIG. 3B, in which inconsistencies (e.g., vertical and / or horizontal inconsistencies) in the Raman-based spectral dataset (e.g., including Raman-based spectral datasets 302a, 302b, and 302c as shown in FIG. 3A) are eliminated or reduced to produce a modified Raman-based spectral dataset (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) with less variability as shown in FIG. 3B.

[0064] The application of the derivative transformation visualized by Figure 3B is further illustrated by the computer program listings of Figures 6A-6C. For example, in the computer program listing embodiment of Figures 6A-6C, in code section 4, the biological classification model configuration includes script executable by first processor 110 of configurable handheld biological analyzer 102, which applies the derivative transformation algorithm as described with respect to Figure 3B herein.

[0065] Figure 3C shows an example visualization 322 of a normalized Raman-based spectral dataset as a normalized version of the modified Raman-based spectral dataset of Figure 3B. For example, Figure 3C may represent one or more subsequent stages in the execution sequence of a spectral pre-processing algorithm. The visualization 322 of Figure 3C includes the same Raman intensity axis 304 and Raman shift axis 306 described herein with respect to Figures 3A and 3B. For example, in one embodiment, the modified Raman-based spectral dataset shown in Figure 3B (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) is aligned across the Raman shift axis 306 by a processor (e.g., first processor 110) to generate the aligned Raman-based spectral dataset (e.g., including Raman-based spectral datasets 322a, 322b, and 322c) shown in Figure 3C. Such alignment applies correction for slight y-axis shifts (i.e., of the Raman intensity axis 304) caused by mismatches / differences between the analytical devices described herein. Application of the alignment algorithm visualized by FIG. 3C is further illustrated by the computer program listings of FIGS. 6A-6C. For example, in the computer program listing embodiment of FIGS. 6A-6C, in code section 6, the biological classification model configuration (e.g., biological classification model configuration 103) includes script executable by the first processor 110 of the configurable handheld biological analyzer 102 to apply a mean centering algorithm to adjust the alignment of the modified Raman-based spectral datasets (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) shown in FIG. 3B to eliminate or reduce spectral mismatches (e.g., vertical and / or horizontal mismatches) between these modified Raman-based spectral datasets. This adjustment results in the aligned Raman-based spectral dataset (eg, including Raman-based spectral datasets 322a, 322b, and 322c) shown in FIG. 3C.

[0066] Additionally or alternatively, in another embodiment, the corrected Raman-based spectral dataset shown in FIG. 3B (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) is normalized across the Raman intensity axis 304 by a processor (e.g., first processor 110) to generate the aligned Raman-based spectral dataset shown in FIG. 3C (e.g., including Raman-based spectral datasets 322a, 322b, and 322c). Such normalization applies a robust normalization algorithm that compensates for intensity-axis variations (i.e., variations in intensity values across the Raman intensity axis 304) caused by inconsistencies / differences between analytical devices described herein. Application of the normalization algorithm visualized by FIG. 3C is further illustrated by the computer program listings of FIGS. 6A-6C. 6A-6C, in code section 5, the biological classification model configuration includes script executable by first processor 110 of configurable handheld biological analyzer 102 to apply a normalization algorithm to normalize the modified Raman-based spectral datasets shown in FIG. 3B (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) to eliminate or reduce spectral inconsistencies (e.g., vertical and / or horizontal inconsistencies) of these modified Raman-based spectral datasets. This normalization results in the normalized Raman-based spectral datasets shown in FIG. 3C (e.g., including Raman-based spectral datasets 322a, 322b, and 322c). Specifically, in the embodiment of Figures 6A-6C, for example, first processor 110 applies a standard normal variate (SNV) algorithm to the corrected Raman-based spectral dataset shown in Figure 3B (e.g., including Raman-based spectral datasets 312a, 312b, and 312c) to generate the aligned Raman-based spectral dataset shown in Figure 3C (e.g., including Raman-based spectral datasets 322a, 322b, and 322c).

[0067] Applying an alignment and / or normalization algorithm (e.g., as described with respect to FIG. 3C ) eliminates or reduces the spectral inconsistencies in the modified Raman-based spectral dataset (e.g., including Raman-based spectral datasets 312 a, 312 b, and 312 c), as shown in FIG. 3B . This is illustrated by a comparison of visualization 312 of FIG. 3B with visualization 322 of FIG. 3C , where the spectral inconsistencies (e.g., vertical and / or horizontal inconsistencies) in the Raman-based spectral dataset (e.g., including Raman-based spectral datasets 312 a, 312 b, and 312 c, as shown in FIG. 3B ) are eliminated or reduced, producing the aligned and / or normalized Raman-based spectral dataset (e.g., including Raman-based spectral datasets 322 a, 322 b, and 322 c) with reduced variability, as shown in FIG. 3C .

[0068] 2, the biological analysis method 200 includes identifying or classifying a biological product type based on a first Raman-based spectral dataset (e.g., the Raman-based spectral dataset visualized and described with respect to FIGS. 3A-3C) using a biological classification model. For example, in various embodiments, once the execution sequence of the spectral pre-processing algorithms is performed (e.g., by the first processor 110), e.g., as described herein with respect to FIGS. 3A-3C and / or 6A-6C, the pre-processed Raman-based dataset, e.g., the aligned and / or normalized Raman-based spectral dataset shown in FIG. 3C (e.g., including Raman-based spectral datasets 322a, 322b, and 322c), can be used by a configurable handheld biological analyzer (e.g., the configurable handheld biological analyzer 102) to identify or classify a biological product (e.g., biological product 140).

[0069] FIG. 5 shows an exemplary visualization 500 of Raman spectra of biological product types (e.g., biological product types 511, 512, and 513). Each of the biological product types (e.g., biological product types 511, 512, and 513) can be identified, classified, or differentiated using a biological classification model (e.g., a biological classification model of biological classification model configuration 103) based on a classification component according to various embodiments disclosed herein. In the embodiment of FIG. 5, each of biological product types 511, 512, and 513 is a different biological product type, including adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513), respectively. Visualization 500 of FIG. 5 includes a Raman intensity axis 504 and a Raman shift axis 506 that are the same as or similar to those described herein with respect to FIGS. 3A and 3B. However, each biological product type 511, 512, and 513 exhibits its own distinct Raman shift axis, where each Raman shift axis exhibits a Raman intensity value between 0 and about 3. Furthermore, Raman shift axis 506 exhibits a Raman intensity value between about 0 and 3000 cm -1 The frequency / wavelength range of

[0070] As shown in FIG. 5, each of the biological product types 511, 512, and 513 are located across the Raman shift axis 506, i.e., in the same or similar Raman spectral range (e.g., 0-3000 cm as shown in FIG. 5). -1 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513). This similar pattern / signature makes it difficult for a typical analytical device to accurately identify, classify, or measure the biological product types Adalimumab DS (biological product type 511), Erenumab DP (biological product type 512), and Romosozumab DP (biological product type 513). A typical analytical device (that does not implement or perform the biological classification model configuration 103 described herein) will typically generate a significant number of Type I (e.g., false positives) and Type II errors (e.g., false negatives) when attempting to identify, measure, or classify these biological product types.

[0071] However, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) loaded with and running a biological classification model configuration described herein (e.g., biological classification model configuration 103) can be used to accurately identify, classify, measure, or differentiate between the biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513). This is shown, for example, in Figure 5, where the biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513) are each identified, classified, and / or measured as distinct from one another by different local features in the Raman spectra (e.g., local features 511c, 512c, and 513c). In the embodiment of FIG. 5, for example, each of local features 511c, 512c, and 513c of biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513), respectively, has a 1000 cm -1 ~1100cm -1 The Raman shift varies across the axis 506 over a range of 1000 cm -1 ~1100cm -1 Each of local features 511c, 512c, and 513c has a different Raman intensity value (having different shapes, peaks, or other different / different relative intensities), specific to each of the biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513), respectively, across a range of 511c, 512c, and 513c. Thus, the different local features (e.g., local features 511c, 512c, and 513c) provide a source of product-specific information that can be used by configurable handheld biological analyzer 102 to identify, classify, or differentiate the biological products described herein.

[0072] Additionally or alternatively, identification or classification is further illustrated with reference to Figure 5, where, for example, biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513) are each identified, classified, and / or measured as distinct from one another by their respective Raman shift axes, i.e., across Raman shift axis 506 (even though the biological products have similar and / or identical Raman spectra). For example, adalimumab DS (biological product type 511) has a first Raman intensity value 511a of about 1.9 (at a Raman shift value of about 2900) and a second Raman intensity value 511z of about 2.25 (at a Raman shift value of about 140). In contrast, erenumab DP (biological product type 512) has a first Raman intensity value 512a (at a Raman shift value of about 2900) of about 2.1 and a second Raman intensity value 512z (at a Raman shift value of about 140) of about 2.5. In further contrast, romosozumab DP (biological product type 513) has a first Raman intensity value 513a (at a Raman shift value of about 2900) of about 1.5 and a second Raman intensity value 513z (at a Raman shift value of about 140) of about 2.05.

[0073] Thus, as shown by visualization 500 in FIG. 5, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) loaded with and running a biological classification model configuration described herein (e.g., biological classification model configuration 103) is highly sensitive to relative differences in Raman intensity values (e.g., on Raman intensity axis 504) between various analyzers and the overall shape of the Raman features (i.e., Raman intensity profile across a range of Raman shift values (Raman shift axis 506)). This is at least in part because a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) loaded with and running a biological classification model configuration (e.g., biological classification model configuration 103) described herein has scan data (a data set of Raman-based spectra) for each of the biological product types adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513) preprocessed with the spectral preprocessing algorithm described herein. Furthermore, the biological classification model used by the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) is further configured to identify the biological product type of the first biological product sample based on the classification component (i.e., to implement a model having a classification component), which also reduces discrepancies, thereby improving the ability of the configurable handheld biological analyzer 102 to identify the biological product type of the first biological product sample.

[0074] In various embodiments, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) identifies, classifies, and / or measures the biological product type of a biological product (e.g., biological product 140), such as adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and romosozumab DP (biological product type 513), based on classification components loaded from a biological classification model configuration (e.g., biological classification model configuration 103). For example, a biological classification model, such as that loaded into configurable handheld biological analyzer 102 via biological classification model configuration 103, may include classification components selected to reduce at least one of (1) the Q residual error of the biological classification model, or (2) a summary value of the fit of the biological classification model, as further described herein in connection with Figures 4A and 4B, respectively.

[0075] As used herein, the term "classification component" may include principal components determined for principal component analysis (PCA). In other embodiments, more generally, classification components may be coefficients or variables of a multivariate model (such as a regression model or a machine learning model). Based on the classification components, the biological classification model is configured to identify a biological product type for a given biological product sample (e.g., selected from biological products 140).

[0076] In some embodiments, the biological classification model may be implemented as a PCA model. Implementing PCA refers to the use of multivariate analysis implemented by the configurable handheld biological analyzer 102 configured with the biological classification model configuration 103 (e.g., as described herein with respect to FIG. 5 ) to identify biological products (e.g., biological products 140), such as therapeutic products / agents, with similar formulations. For example, biological or pharmaceutical products typically involve high-dimensional data. High-dimensional data may include multiple features, such as the expression of many genes measured on a given sample (e.g., a sample from a scan of the biological product 140). PCA provides a technique used by the configurable handheld biological analyzer 102 to simplify the complexity of high-dimensional data (e.g., a dataset of Raman spectra) while simultaneously retaining trends and patterns useful for prediction and / or identification purposes (e.g., biological product identification as described herein). For example, applying PCA may include converting (e.g., by the first processor 110) a dataset (e.g., a dataset of Raman-based spectra) to a lower dimensionality. The lower-dimensional transformed dataset provides a summary or simplification of the original dataset. The transformed dataset subsequently reduces computational costs when manipulated by a configurable handheld biological analyzer described herein (e.g., configurable handheld biological analyzer 102). Furthermore, the error rates described herein may also be reduced by implementing PCA, thereby eliminating the need to apply test corrections to higher-dimensional data when testing each feature to associate it with a particular outcome.

[0077] Furthermore, when implemented by the configurable handheld biological analyzer 102, PCA reduces data complexity by geometrically projecting the data into lower dimensions called principal components (PCs) and by using a limited number of PCs to target the best summary of the data, and thus the PCs. The first PC is selected to minimize the total distance between the data and its projection onto the PC. Any second (subsequent) PC is similarly selected with the additional requirement that it be uncorrelated with all previous PCs.

[0078] PCA is an unsupervised learning method similar to clustering. That is, it discovers trends or patterns without reference to prior knowledge regarding whether samples came from different sources, such as different configurable handheld biological analyzers (e.g., configurable handheld biological analyzers 102, 112, 114, and / or 116). For example, in some embodiments, the classification component of the biological classification model may be the first principal component of the PCA model. In such embodiments, the first principal component may be determined by the first processor 110 based on a singular value decomposition (SVD) analysis. The use of the first principal component by the configurable handheld biological analyzer 102 limits or reduces the amount of analyzer variability compensated for by the biological classification model. In some embodiments, the first principal component (PC) may be the only principal component. In other embodiments, the biological classification model may include a second classification component, where the biological classification model is configured to identify a biological product type of a given biological product sample (e.g., biological product 140) based on multiple classification components (e.g., a first classification component and a second classification component).

[0079] In the computer program listing embodiment of FIGS. 6A-6C, in code section 7, a biological classification model configuration (e.g., biological classification model configuration 103) defines a set of PCA predictions specified for that biological classification model. Code section 7 also includes summary statistics of the fit (e.g., Hotelling T 2 The script in code section 7 is also provided to define calculations for the Q residual / values and the Hotelling T residual / values, as described herein, for example, with respect to FIGS. 4A and 4B. 2 Based on the value, the biological product (eg, biological product 140) can be identified or classified.

[0080] 4A shows an example visualization 400 of the Q-residual error of a biological classification model. 2 axis 406. In general, the Q residual error and the Hotelling T 2 The values are summary statistics that can be used to describe how well a model (e.g., a biological classification model of biological classification model configuration 103) describes a given biological product sample (e.g., taken from a scan of biological product 140). Figure 4A shows the Q residual error and Hotelling T residual error for a number of handheld biological analyzers. 2 Plot the values. Generally, a Q residual error of zero (0) and a Hotelling T 2 A handheld biological analyzer with a value represents an error-free scan of the product.

[0081] The handheld biological analyzers include handheld biological analyzers in biological analyzer groups 411n, 412m1, 412m2, and 413n. Analyzer group 411n represents analyzers that scanned a first biological product type, adalimumab DS. Analyzer groups 412m1 and 412m2 each represent analyzers that scanned a second biological product type, erenumab DP. Analyzer group 413n represents analyzers that scanned a third biological product type, romosozumab DP. Analyzer groups 412m1 and 412m2 include configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116) configured and enhanced with the biological classification model configurations (e.g., biological classification model configuration 103) and respective biological classification models described herein. Analytical device groups 411n and 413n include typical biological analytical devices that are not configured with biological classification models or biological classification models.

[0082] Analyzer groups 411n and 413n serve as control groups that, when compared to analyzer groups 412m1 and 412m2, demonstrate improvement over a typical analyzer, for example, through a reduction in error (e.g., along Q-residual error axis 404) of the configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116). Specifically, the Q-residual (e.g., on Q-residual error axis 404) provides a lack-of-fit statistic calculated as the sum of squares for each product sample. The Q-residual represents the amount of variability remaining in each sample after projection through a given model (e.g., a biological classification model described herein). More generally, as illustrated by the embodiment of FIG. 4A, the Q-residual value (along Q-residual error axis 404) serves as a discriminatory statistic. Q-residual is a measure of what is "left over," i.e., what is not explained, by a given biological classification model. For example, in one embodiment where the biological classification model is implemented as a PCA model (e.g., where spectra are projected onto the first principal component), the values in Figure 4A represent what is left over (residual) after the scanned data (e.g., for biological analyzer groups 411n, 412m1, 412m2, and / or 413n) are projected by the first principal component.

[0083] In various embodiments, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) includes a biological classification model (e.g., of biological classification model configuration 103) configured to identify or classify a biological product type of a biological product sample (e.g., taken from biological product 140) based on a classification component when the Q-residual error meets a threshold. In some embodiments, the biological classification model implemented or executed by, for example, first processor 110 of configurable handheld biological analyzer 102 outputs a pass / fail decision based on a threshold. For example, in the embodiment of FIG. 4A , a threshold of “1” across Q-residual error axis 404 is selected as pass / fail determinant threshold 405. In such an embodiment, a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) implementing the biological classification model will identify or classify (i.e., "pass") these biological products using scanned data (e.g., Raman spectral data sets) that are within (i.e., below) a threshold value of 1 across the Q residual error axis 404. Otherwise, the biological analyzer (e.g., configurable handheld biological analyzer 102) implementing the biological classification model will not identify or classify (i.e., "fail") these biological products.

[0084] 4A , analytical device groups 412m1 and 412m2 include configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116) configured and enhanced with a biological classification model configuration (e.g., biological classification model configuration 103) and respective biological classification model described herein. The configurable handheld biological analyzers of analytical device groups 412m1 and 412m2 correctly identified or classified (i.e., "passed") the biological product (i.e., erenumab DP), where the associated scanned data (e.g., Raman spectral dataset), when preprocessed with a spectral preprocessing algorithm described herein, falls within (i.e., below) a threshold value of 1, as shown by visualization 400.

[0085] Thus, a biological classification model of a configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) may include classification components selected to reduce the Q-residual error of the biological classification model. In this manner, the biological classification model is configured to identify the biological product type of a given biological product sample based on the classification components. Generally, Q-residuals are most appropriate for use with single-specification biological products when lot-to-lot variability is the primary source of discrepancy between analyzers. Thus, as shown by FIG. 4A, Q-residuals can be used as a discriminative statistic to determine models (e.g., biological classification models described herein) that are tolerant to analyzer-to-analyzer variability.

[0086] Figure 4B shows the summary values of the fit of the biological classification model (e.g., Hotelling T 2 4 shows an exemplary visualization 450 of the Hotelling T 2 The values represent a measure of variability for each sample within a model (e.g., a biological classification model). 2 The values indicate how far each sample is from the "midpoint" of the model (a value of 0). In other words, the Hotelling T 2 The value is an indicator of the distance from the model midpoint. Distance from the midpoint can often occur due to variations between analyzers. Hotelling T 2 The use of Q values is advantageous in identifying biological products using multiple standards, where varying concentrations of active ingredients, excipients, etc., cause variability in the Raman spectra that is greater than lot-to-lot variability (as discussed herein above with respect to Q residuals with reference to Figure 4A).

[0087] In the embodiment of FIG. 4B, the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) calculates a summary value of the fit (e.g., Hotelling T 2 4B includes a biological classification model (e.g., of biological classification model configuration 103) configured to identify or classify a biological product type of a biological product sample (e.g., taken from biological product 140) based on the classification component when the Q residual error axis 404 and Hotelling T are met a threshold. 2 4B includes axis 406. Analyzer group 452m represents analyzers that scanned a first biological product type, Denosumab DP (with two standards). Analyzer group 454m represents analyzers that scanned a second biological product type, Denosumab DS (with one standard). Analyzer group 462n represents analyzers that scanned a third biological product type, Enbrel DP. In the embodiment of FIG. 4B, Hotelling T 2 A threshold of "1" across axis 406 is selected as the pass / fail determinant threshold 407. In such an embodiment, the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) implementing the biological classification model is based on the Hotelling T 2 Scanned data (e.g., Raman spectral data sets) that fall within (i.e., below) a threshold value of one on axis 406 are used to identify or classify (i.e., "pass") these biological products. Otherwise, the biological analyzer (e.g., configurable handheld biological analyzer 102) implementing the biological classification model will not identify or classify (i.e., "fail") these biological products.

[0088] In the embodiment of FIG. 4B, the biological classification model of the configurable handheld biological analyzer (e.g., configurable handheld biological analyzer 102) is calculated based on a summary value of the biological classification model fit (e.g., Hotelling T 2 The biological classification model may include classification components selected to reduce the overall biological product type (e.g., a value). In this manner, the biological classification model is configured to identify a biological product type for a given biological product sample based on the classification components. For example, analytical device groups 452m and 454m include configurable handheld biological analyzers (e.g., any of configurable handheld biological analyzers 102, 112, 114, and / or 116) configured and powered with the biological classification model configurations (e.g., biological classification model configuration 103) and respective biological classification models described herein. The configurable handheld biological analyzers of analytical device groups 412m1 and 412m2 correctly identify or classify (i.e., "pass") the biological products (i.e., denosumab DP and DS), where the associated scanned data (e.g., Raman spectral datasets), when preprocessed with the spectral preprocessing algorithms described herein, falls within (i.e., falls below) a threshold value of 1, as shown by visualization 450. In contrast, analytical device group 462n may represent analytical devices that are not configured with the biological classification model configurations described herein.

[0089] As shown by each of FIGS. 4A and 4B, the Q residual error (e.g., on the Q residual error axis 404) and / or the Hotelling T 2 Each of the values can be used alone or together to identify or classify a biological product. That is, the configurable handheld biological analyzer 102 can be configured to select or implement a classification component to reduce one or both of: (1) the Q residual error of the biological classification model; and / or (2) the summary value of the fit of the biological classification model.

[0090] As described herein with reference to Figures 2, 3A, 3B, 3C, 4A, 4B, and 5, a biological classification model can be configured to identify, classify, measure, or otherwise distinguish a given biological product sample having a given biological product type (e.g., adalimumab DS (biological product type 511)) from a different or second biological product sample having a different or second biological product type (e.g., erenumab DP (biological product type 512)) based on the classification component. For example, as described with reference to Figures 4A, 4B, and 5, the configurable handheld biological analyzer 102 can distinguish between a first biological product type (e.g., adalimumab DS (biological product type 511)) and a different biological product type (e.g., erenumab DP (biological product type 512)). For example, as described herein, the configurable handheld biological analyzer 102, once configured with the biological classification model configuration 103, can perform a spectral pre-processing algorithm (e.g., as described in Figures 3A-3C herein) on a dataset of Raman-based spectra received by the first scanner 106. Once the dataset of Raman-based spectra has been pre-processed by the spectral pre-processing algorithm, the configurable handheld biological analyzer 102 can identify or classify biological products based on Q residuals and / or Hotelling T2 values (e.g., as described herein with respect to Figures 4A and 4B).

[0091] During development or manufacturing of a biological product, such as biological product 140 having a given biological product type, for example, any of the adalimumab DS (biological product type 511), erenumab DP (biological product type 512), and / or romosozumab DP (biological product type 513) described herein, the biological product type may be identified by the configurable handheld biological analyzer 102 (e.g., by the first processor 110) executing a biological classification model and / or a spectral pre-processing algorithm. However, it should be understood that these biological product types are examples only, and that other biological product types or biological products may be identified, classified, measured, or differentiated in the same or similar manner as described with respect to various embodiments herein.

[0092] Aspects of the Disclosure 1. A configurable handheld biological analyzer for identifying biological products based on Raman spectroscopy, comprising: a first housing adapted for handheld operation; a first scanner carried by the first housing; a first processor communicatively coupled to the first scanner; and a first computer memory communicatively coupled to the first processor, the first computer memory configured to load a biological classification model configuration, the biological classification model configuration including a biological classification model, the biological classification model configured to run on the first processor, the first processor configured to (1) receive a first Raman-based spectral dataset defining a first biological product sample scanned by the first scanner; and (2) use the biological classification model to generate the first Raman-based spectral dataset. and a biological classification model configured to identify a biological product type of the first biological product sample based on a Raman-based spectral dataset, the biological classification model configuration further comprising a spectral pre-processing algorithm, the first processor configured to execute the spectral pre-processing algorithm when the first Raman-based spectral dataset is received by the first processor to reduce spectral discrepancies in the first Raman-based spectral dataset, the biological classification model including a classification component selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model configured to identify a biological product type of the first biological product sample based on the classification component.

[0093] 2. The configurable handheld biological analyzer of aspect 1, wherein the biological classification model configuration is electronically transferable to a second configurable handheld biological analyzer, the second configurable handheld biological analyzer including a second housing adapted for handheld operation, a second scanner coupled to the second housing, a second processor communicatively coupled to the second scanner, and a second computer memory communicatively coupled to the second processor, the second computer memory configured to load the biological classification model configuration, the biological classification model configuration including a biological classification model, the biological classification model configured to execute on the second processor, the second processor configured to (1) receive a second Raman-based spectral dataset defining a second biological product sample scanned by the second scanner, and (2) identify a biological product type based on the second Raman-based spectral dataset using the biological classification model, wherein the second biological product sample is a novel sample of the biological product type.

[0094] 3. The configurable handheld biological analyzer of any of aspects 1 or 2, wherein the spectral mismatch is an inter-analyzer spectral mismatch between the first Raman-based spectral dataset and one or more other Raman-based spectral datasets of one or more corresponding other handheld biological analyzers, each of the one or more other Raman-based spectral datasets representing a biological product type, and wherein the spectral pre-processing algorithm is configured to reduce the inter-analyzer spectral mismatch between the first Raman-based spectral dataset and the one or more other Raman-based spectral datasets.

[0095] 4. The configurable handheld biological analyzer of aspect 3, wherein the spectral pre-processing algorithm includes applying a derivative transform to the first Raman-based spectral dataset to generate a modified Raman-based spectral dataset; aligning the modified Raman-based spectral dataset across a Raman shift axis; and normalizing the modified Raman-based spectral dataset across a Raman intensity axis.

[0096] 5. The configurable handheld biological analyzer of aspect 4, wherein the derivative transformation includes determining a Raman weighted average of a consecutive group of 11 to 15 Raman shift values across the Raman shift axis, and determining a corresponding derivative of these Raman weighted averages across the Raman shift axis.

[0097] 6. The configurable handheld biological analysis device of any of aspects 1-5, wherein the classification component is selected to reduce both (1) the Q residual error of the biological classification model and (2) the summary value of the fit of the biological classification model.

[0098] 7. The configurable handheld biological analyzer of any of aspects 1-6, wherein the biological classification model further includes a second classification component, and the biological classification model is configured to identify a biological product type of the first biological product sample based on the classification component and the second classification component.

[0099] 8. The configurable handheld biological analyzer of any of aspects 1-7, wherein the biological classification model is implemented as a principal component analysis (PCA) model.

[0100] 9. The configurable handheld biological analyzer of aspect 8, wherein the classification component is the first major component of the PCA model.

[0101] 10. The configurable handheld biological analysis device of any of aspects 1-9, wherein the computer memory is configured to load a new biological classification model, the new biological classification model including an updated classification component.

[0102] 11. The configurable handheld biological analyzer of any of aspects 1-10, wherein the biological classification model configuration is implemented in an Extensible Markup Language (XML) format.

[0103] 12. The configurable handheld biological analytical device of any of aspects 1-11, wherein the biological product type is a therapeutic product.

[0104] 13. The configurable handheld biological analytical device of any of aspects 1-12, wherein the biological product type is identified by a biological classification model during production of a biological product having the biological product type.

[0105] 14. The configurable handheld biological analyzer of any of aspects 1-13, wherein the biological classification model is configured to distinguish a first biological product sample having a biological product type from a different biological product sample having a different biological product type based on the classification component.

[0106] 15. The configurable handheld biological analyzer of aspect 14, wherein the biological product type and the different biological product type each have distinct localized features within the same or similar Raman spectral range.

[0107] 16. The configurable handheld biological analyzer of any of aspects 1-15, wherein the biological classification model is configured to identify a biological product type of the first biological product sample based on the classification component when a Q residual error or a summary fit value meets a threshold.

[0108] 17. The configurable handheld biological analysis device of aspect 16, wherein the biological classification model outputs a pass / fail decision based on a threshold value.

[0109] 18. The configurable handheld biological analytical device of any of aspects 1-17, wherein the biological classification model is generated by a remote processor that is remote to the configurable handheld biological analytical device.

[0110] 19. A biological analytical method for identifying a biological product based on Raman spectroscopy, comprising: loading a biological classification model configuration including a biological classification model into a first computer memory of a first configurable handheld biological analytical device having a first processor and a first scanner; receiving a first Raman-based spectral dataset defining a first biological product sample scanned by the first scanner using the biological classification model; executing a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the first Raman-based spectral dataset; and identifying a biological product type based on the first Raman-based spectral dataset using the biological classification model, wherein the biological classification model includes a classification component selected to reduce at least one of (1) a Q residual error of the biological classification model or (2) a summary value of fit of the biological classification model, and the biological classification model is configured to identify a biological product type of the first biological product sample based on the classification component.

[0111] 20. The biological analytical method of embodiment 19, wherein the biological classification model configuration is electronically transferable to a second configurable handheld biological analytical device, and the biological analytical method further includes loading the biological classification model configuration, including the biological classification model, into a second computer memory of the second configurable handheld biological analytical device having a second processor and a second scanner; receiving a second Raman-based spectral dataset that defines a second biological product sample scanned by the second scanner according to the biological classification model; executing a spectral pre-processing algorithm of the biological classification model to reduce a second spectral discrepancy in the second Raman-based spectral dataset; and identifying a biological product type based on the second Raman-based spectral dataset using the biological classification model, wherein the second biological product sample is a novel sample of the biological product type.

[0112] 21. The biological analytical method of aspect 19 or 20, wherein the spectral mismatch is an inter-analytical device spectral mismatch between the first Raman-based spectral dataset and one or more other Raman-based spectral datasets of one or more corresponding other handheld biological analytical devices, each of the one or more other Raman-based spectral datasets representing a biological product type, and wherein the spectral pre-processing algorithm is configured to reduce the inter-analytical device spectral mismatch between the first Raman-based spectral dataset and the one or more other Raman-based spectral datasets.

[0113] 22. The biological analytical method of aspect 21, wherein the spectral pre-processing algorithm includes applying a derivative transform to the first dataset of Raman-based spectra to generate a dataset of modified Raman-based spectra; aligning the dataset of modified Raman-based spectra across a Raman shift axis; and normalizing the dataset of modified Raman-based spectra across a Raman intensity axis.

[0114] 23. The biological analytical method of embodiment 22, wherein the derivative transformation comprises determining Raman weighted averages of consecutive groups of 11 to 15 Raman shift values across the Raman shift axis, and determining corresponding derivatives of these Raman weighted averages across the Raman shift axis.

[0115] 24. The biological analysis method of any one or more of aspects 19-23, wherein the classification component is selected to reduce both (1) the Q residual error of the biological classification model and (2) the summary value of the fit of the biological classification model.

[0116] 25. The biological analytical method of any one or more of aspects 19-24, wherein the biological classification model further comprises a second classification component, and the biological classification model is configured to identify a biological product type of the first biological product sample based on the classification component and the second classification component.

[0117] 26. The biological analytical method according to any one or more of aspects 19 to 25, wherein the biological classification model is implemented as a principal component analysis (PCA) model.

[0118] 27. The biological analytical method according to aspect 26, wherein the classification component is the first principal component of the PCA model.

[0119] 28. The biological analysis method of any one or more of aspects 19-27, wherein the first and / or second computer memory is configured to load a new biological classification model, and the new biological classification model includes an updated classification component.

[0120] 29. The biological analytical method according to any one or more of aspects 19 to 28, wherein the biological classification model configuration is implemented in an Extensible Markup Language (XML) format.

[0121] 30. The biological analytical method according to any one or more of aspects 19 to 29, wherein the biological product type is a therapeutic product.

[0122] 31. The biological analytical method according to any one or more of aspects 19 to 30, wherein the biological product type is identified by a biological classification model during production of a biological product having the biological product type.

[0123] 32. The biological analytical method of any one or more of aspects 19-31, wherein the biological classification model is configured to distinguish a first biological product sample having a biological product type from a different biological product sample having a different biological product type based on the classification component.

[0124] 33. The biological analytical method according to aspect 32, wherein the biological product type and the different biological product type each have the same or similar Raman spectral ranges.

[0125] 34. The biological analytical method of any one or more of aspects 19-33, wherein the biological classification model is configured to identify a biological product type of the first biological product sample based on the classification component when the Q residual error or the summary value of the fit meets a threshold.

[0126] 35. The biological analytical method of embodiment 34, wherein the biological classification model outputs a pass / fail decision based on a threshold value.

[0127] 36. The biological analytical method of any one or more of aspects 19-35, wherein the biological classification model is generated by a remote processor that is remote to the configurable handheld biological analytical device.

[0128] 37. A tangible, non-transitory computer-readable medium storing instructions for identifying biological products based on Raman spectroscopy that, when executed by one or more processors of a configurable handheld biological analyzer, cause the one or more processors of the configurable handheld biological analyzer to: load a biological classification model configuration into a computer memory of a configurable handheld biological analyzer having a scanner, the biological classification model configuration including a biological classification model; receive a Raman-based spectral dataset defining a biological product sample scanned by the scanner using the biological classification model; execute a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the Raman-based spectral dataset; and identify a biological product type based on the Raman-based spectral dataset using the biological classification model, the biological classification model including classification components selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model is configured to identify a biological product type of the biological product sample based on the classification components.

[0129] The above-described aspects of the present disclosure are examples only and are not intended to limit the scope of the present disclosure.

[0130] Additional Examples The following additional examples provide further support in accordance with various embodiments described herein. Specifically, the following additional examples demonstrate Raman spectroscopy for rapid identity (ID) verification of biotherapeutic protein products in solution. This example demonstrates the unique combination of Raman features associated with both therapeutic agents and excipients as the basis for product differentiation. The product ID methods (e.g., bioanalytical methods) described herein involve acquiring Raman spectra of target products on multiple Raman analyzers (e.g., configurable handheld bioanalytical devices described herein). The spectra may then be subjected to principal component analysis (PCA) to define product-specific models (e.g., biological classification models) that serve as the basis for product ID determination on the configurable handheld bioanalytical devices, and dimension reduction using the bioanalytical methods to identify biological products based on the Raman spectroscopy described herein. The product-specific models (e.g., biological classification models) may be transferred to individual instruments (e.g., configurable handheld bioanalytical devices) where they have been validated for product testing. These can be used for a variety of purposes, including quality control, point-of-delivery quality assurance, and manufacturing. Such analytical devices and methods can be used across a variety of Raman instruments (e.g., configurable handheld biological analyzers) from a variety of manufacturers. Thus, additional examples further demonstrate that the Raman ID analytical devices and methods (e.g., configurable handheld biological analyzers and related methods) described herein offer a variety of applications and testing for solution-based protein products in the biopharmaceutical industry.

[0131] Additional Examples - Materials Drug substances and drug products corresponding to over 28 individual product specifications were analyzed during the development and testing of the configurable handheld bioanalytical device and associated methods described herein. Table 1 shows the active pharmaceutical ingredient (API) concentration and molecular class for 14 product specifications representing a set of late-stage and commercial product specifications. Product solutions were transferred to 4 mL glass vials to serve as sample cells for Raman spectral acquisition. Table 1 provides general properties of the products evaluated, either as targets for ID methods (e.g., bioanalytical methods) or as specificity challenges described herein. For clarity in Table 1, each product is labeled with a letter code. Products with the same character letter but different numbers (e.g., A1 and A2) represent products with the same active ingredient, which may differ in protein concentration and / or formulation. The listed materials can be used to manufacture the formulations. It will be understood that some formulations may be identified, for example, by trade names referenced herein.

[0132] [Table 1]

[0133] Additional Examples—Raman Instrumentation (e.g., Configurable Handheld Bioanalyzers) and Measurements For additional examples, Raman spectra were measured using a configurable handheld biological analyzer described herein. For example, in certain embodiments, the configurable handheld biological analyzer may be a Raman-based handheld analyzer, such as the TruScan™ RM Handheld Raman Analyzer offered by Thermo Fisher Scientific Inc. In such embodiments, the configurable handheld biological analyzer may implement the TruTools™ chemometrics software package. However, it should be understood that other brands or types of Raman analyzers using additional and / or different software packages may be used in accordance with the disclosure herein. In some embodiments, for sample integration, the configurable handheld biological analyzer may be configured with a 785 nm grating-stabilized laser source (maximum output power 250 mW) coupled with focusing optics (e.g., 0.33 NA, 18 mm working distance, spot >0.2 mm). For additional examples, a product solution contained in a glass vial was secured in front of the focusing optics using a vial adapter on the configurable handheld biological analyzer. All spectra were collected using the same spectral acquisition settings, e.g., laser power = 250 mW, integration time = 1000 ms, and number of spectral co-additions = 70 (although other settings may be used). For additional examples, product spectra were collected over a period of time using three different configurable handheld biological analyzers (hereinafter referred to as configurable handheld biological analyzers 1-3) and / or equipment dedicated to the configuration and / or development of biological analytical methods for identifying biological products based on Raman spectroscopy as described herein. It should be understood that additional or fewer analyzers, using the same or different settings, may be used to set up, configure, or initialize the configurable handheld biological analyzers and associated biological analytical methods described herein.

[0134] Additional Example - Development of a Multivariate Raman ID Bioanalytical Method For example, a Raman spectral model (e.g., a biological classification model) based on principal component analysis (PCA) can be generated, developed, or loaded as described herein. For example, in some embodiments, a Raman spectral model (e.g., a biological classification model) can be generated, developed, or loaded using SOLO software (Solo+Model_Exporter version 8.2.1; Eigenvector Research, Inc.) equipped with the Model Exporter add-on. However, it should be understood that other software may be used to generate, develop, or load a Raman spectral model (e.g., a biological classification model). Spectra used for model building may typically be collected as replicate scans on two or more separate lots of material using a configurable handheld biological analyzer (e.g., three configurable handheld biological analyzers). Spectra are typically acquired over multiple days to include instrument drift. In some embodiments, the spectral range may be reduced to remove background variations resulting from detector noise >1800 cm and optics that filter out Rayleigh lines <400 cm before incorporation into a model (e.g., a biological classification model). Spectra may be further preprocessed and mean-centered as described herein for each model. Models may be further refined by cross-validation using a random subset procedure by reference to the Raman spectra of the target and challenge products shown in Table 1.

[0135] The biological classification model configuration (e.g., PCA model configuration), along with Raman spectral acquisition parameters, may be configured or loaded into a configurable handheld biological analyzer and / or may use the biological analysis methods for identifying biological products based on Raman spectroscopy described herein. Acceptance (e.g., pass / fail) criteria for each method may also be specified. As described herein, the pass / fail criteria are based on the reduced Hotelling T (reduced Hotelling T), two summary statistics that generally describe how well a Raman spectrum is described by a biological classification model (e.g., PCA model). 2 (T r 2 ) and Q residual (Q r ) threshold. Equations (1)-(4) below provide exemplary user-selectable decision logic options for positive identification or decision (e.g., pass / fail criteria) by a biological classification model (e.g., a PCA model).

number

[0136] In the example equation above, we calculate the Hotelling T by dividing the original value by the corresponding confidence interval, thereby setting the upper bound to a value of 1. 2 The Q residual values are normalized (i.e., reduced T r 2 and Q r ).

[0137] Additional Examples - Configurable Handheld Biological Analytical Device and Transfer Test Method In connection with additional examples, demonstration of the performance of the configurable handheld biological analyzers and associated methods described herein with respect to the five product-specific models (e.g., biological classification models) described herein with respect to Figures 8A-8E was performed using a small battery of analyzers (e.g., 15 configurable handheld biological analyzers) that were not used in the development of the configurable handheld biological analyzers and associated methods described herein, i.e., not previously configured or loaded with the biological classification model configurations described herein. Product ID methods (e.g., biological analysis methods for identifying biological products based on Raman spectroscopy) were prepared on configurable handheld biological analyzers 1-3, and four tests for a single product standard (e.g., Q1, Q2, A1, and A2) and one test suitable for identifying three similar standards of the same protein product (e.g., B1, B2, and B3) were performed. Each test involved using target product spectra acquired on 15 additional instruments (analyzers 4-18), each varying in age and performance. Model specificity was also measured by evaluating the closest specific challenge product and formulation buffer (i.e., no protein). Raman spectra of samples were acquired using the same collection parameters (i.e., laser power, acquisition time, number of co-additions) used for model building. Raman spectra were acquired in replicates over various days, resulting in approximately 250 spectra per product sample. Spectra acquired during testing were evaluated against each of the five PCA models (e.g., in Eigenvector Solo and Model_Exporter software) to assess the likelihood of false-positive (i.e., misidentification of the challenge product as the target) and false-negative (i.e., incorrect rejection of the target product by the model) results.

[0138] During testing of additional examples, for example, there were no instances of false positive results in any of the five models and associated tests described with respect to Figures 8A-8E. r or T r 2 The values were greater for analyzers 4 through 18 than for the instrument used to develop the model. As an extension of this observation, the ability of a biological classification model (e.g., a PCA model) to consistently reject a given challenge product can be reliably estimated based solely on Raman spectra obtained during method development. Figure 7 shows an exemplary visualization 700 of reduced Q residual error 704 according to various embodiments described herein. Specifically, Figure 7 provides an exemplary plot of reduced Q residual error values 700 for product A1 of Table 1, treated as the challenge product sample, evaluated against a biological classification model (e.g., a PCA model) for product A2 of Table 1. Linear indices 706 are provided for the Raman spectral indices in the dataset and are not necessarily associated with the sample.

[0139] The individual points in Figure 7 are differentiated based on whether the corresponding Raman spectrum was acquired on the analytical instrument used to develop the model (702) or was used solely for testing (703). r The Q values (i.e., linear exponent values of approximately 250-270) were abnormally high due to known instrument performance issues, which are discussed herein below. Nevertheless, even excluding measurements made on analyzer 8, the Q values of analyzers 4-18 were significantly higher than those of analyzers 4-18. r The values were not normally distributed based on the rejection of the Shapiro-Wilk null hypothesis (p=0.0013). For this data set, a median Q of 3.02 for analyzers 4-18 was obtained. r The value is a median Q of 2.53 for the developed instrument. r The mean mean variance was significantly greater than the mean mean variance (Mann-Whitney U test p-value <0.0001). No false positives were observed. However, there were 33 false negative predictions that should have been positive identifications across 1540 total measurements, which equates to only approximately 2% false negatives, a small fraction of the total number of analyses.

[0140] 8A-8E show the summary statistics Q for each of the target products evaluated against its corresponding biological classification model (e.g., PCA model). r or T r 2 8A-8D show analysis of various analytical devices 802 (i.e., configurable handheld biological analyzers 1-3 and analyzers 4-18) by plotting the Q residuals of the target products evaluated on the various analytical devices 802. For clarity, the validation results in FIGS. 8A-8E are organized according to the analytical device number. FIGS. 8A-8D show exemplary visualizations 800, 810, 820, and 830, respectively, of the reduced Q residuals of target products (e.g., from Table 1) evaluated on 18 different configurable handheld biological analyzers (configurable handheld biological analyzers 1-3 and analyzers 4-18) according to various embodiments described herein. Specifically, the visualizations in FIGS. 8A-8D are presented as scatter plots showing the spread of the reduced Q residuals of the target products for each method evaluated on analytical devices 1-18. FIG. 8A shows the spread of the reduced Q residuals for target product A1 of Table 1. FIG. 8B shows the spread of the reduced Q residuals for target product A2 of Table 1. Figure 8C shows the spread of reduced Q residuals for target product Q1 in Table 1. Figure 8D shows the spread of reduced Q residuals for target product Q2 in Table 1. In each of Figures 8A-8D, the horizontal dashed lines in each graph (i.e., 805, 815, 825, and 835, respectively) represent pass / fail criteria or thresholds, such that values above 1 result in a failing result (i.e., a false negative). Each linear index (e.g., 806, 816, 826, and 836, respectively) is provided for a Raman spectral index in the data set and is not necessarily associated with a sample.

[0141] FIG. 8E shows reduced fit summary values (e.g., Hotelling T) of target products (e.g., B1, B2, and / or B3) evaluated for 18 different configurable handheld biological analyzers 802 (Configurable Handheld Biological Analyser 1-3 and Analyser 4-18) according to various embodiments described herein. 2 ) is shown. The horizontal dashed line 845 represents a pass / fail criterion or threshold, such that a value above 1 results in a failing result (i.e., a false negative). Linear indices 846 are provided for the Raman spectral indices in the dataset and are not necessarily associated with the sample.

[0142] For each of Figures 8A-8E, there were no false negative determinations on analyzers 10-16, and 18. In fact, in most cases the summary statistic was <0.6, suggesting that the likelihood of a false negative on any of these instruments was extremely low. There were 33 erroneous results isolated to the remaining three analyzers (8, 9, 17), each of which had identifiable hardware-based and / or instrument-specific performance issues. Analyzer 8, an early pilot instrument, produced the highest number of false negatives. For Method A1, the 20 / 20 spectrum failed Q r However, the other four methods only had a total of three false negatives, suggesting that the different performance for Method A1 is likely related to the weak Raman scattering signal for this product due to its low protein concentration (10 mg / mL) and weak excipient bands. Nevertheless, inspection of the analyzer 8 residuals revealed a peak at approximately 1300 cm -1 The results revealed extensive characterization, centered on the Raman bands (data not shown). An early test-built instrument, analyzer 8, had different optical components than the production analyzers (1-7 and 9-18) that resulted in observable Raman bands, which was believed to be responsible for the high rejection rate. Significant instrument performance issues were also noted for the remaining analyzers (9 and 17). The original wavenumber calibration for analyzer 9 was approximately 3 cm off the manufacturer's specification. -1 For analyzer 17, further investigation identified a previously unknown laser power / stability issue. Despite these known issues, the true positive rates for all five models in Figures 8A–8E exceeded 85% on both analyzers, providing evidence that biological classification models (e.g., PCA) can tolerate even some degradation in instrument performance. Procedural mechanisms designed to ensure instrument fit-for-use (e.g., installation and operational qualification, periodic preventive maintenance) are already described in Good Manufacturing Practice (GMP) testing for biopharmaceuticals. However, the fact that the laser power issue for analyzer 17 was not known prior to testing highlights the value of rigorous evaluation of instrument performance metrics to ensure the long-term performance of spectrometers and multivariate models. However, as demonstrated above, the configurable handheld biological analytical devices and associated biological analytical methods described herein for identifying biological products based on Raman spectroscopy are robust and fault-tolerant as described herein and remain operable and usable despite instrument hardware-based and / or instrument-specific performance issues.

[0143] Additional explanation The above description herein describes various devices, assemblies, components, subsystems, and methods of use related to drug delivery devices. The devices, assemblies, components, subsystems, methods, or drug delivery devices may further include or be used in conjunction with drugs, including, but not limited to, the drugs identified below and their generic and biosimilar equivalents. As used herein, the term drug may be used interchangeably with other similar terms and may refer to any type of pharmaceutical or therapeutic material, including traditional and non-traditional medicines, nutraceuticals, supplements, biologics, biologically active agents and compositions, large molecules, biosimilars, bioequivalents, therapeutic antibodies, polypeptides, proteins, small molecules, and generic drugs. Non-therapeutic injectable materials are also encompassed. Drugs may be in liquid form, lyophilized form, or reconstituted from lyophilized form. The following list of exemplary drugs should not be considered exhaustive or limiting.

[0144] The drug is contained in a reservoir. In some cases, the reservoir is a primary container that is either filled or pre-filled with the drug for treatment. The primary container can be a vial, cartridge, or pre-filled syringe.

[0145] In some embodiments, the reservoir of the drug delivery device may be loaded with or the device may be used in conjunction with a colony-stimulating factor, such as granulocyte colony-stimulating factor (G-CSF). Such G-CSF agents include, but are not limited to, Neulasta® (pegfilgrastim, PEGylated filgastim, PEGylated G-CSF, PEGylated hu-Met-G-CSF) and Neupogen® (filgrastim, G-CSF, hu-Met-G-CSF).

[0146] In other embodiments, the drug delivery device may contain or be used in conjunction with an erythropoiesis-stimulating agent (ESA), which may be in liquid or lyophilized form. An ESA is any molecule that stimulates red blood cell production. In some embodiments, the ESA is an erythropoiesis-stimulating protein. As used herein, "erythropoiesis-stimulating protein" refers to any protein that directly or indirectly activates the erythropoietin receptor, for example, by binding to the receptor and causing it to dimerize. Erythropoiesis-stimulating proteins include erythropoietin and variants, analogs, or derivatives thereof that bind to and activate the erythropoietin receptor, antibodies that bind to and activate the erythropoietin receptor, or peptides that bind to and activate the erythropoietin receptor. Erythropoiesis-stimulating proteins include Epogen® (epoetin alfa), Aranesp® (darbepoetin alfa), Dynepo® (epoetin delta), Mircera® (methoxypolyethylene glycol-epoetin beta), Hematide®, MRK-2578, INS-22, Retacrit® (epoetin zeta), Neorecormon® (epoetin beta), Silapo® (epoetin zeta), Binocrit® (epoetin alfa), Epoetin alpha, epoetin beta, epoetin iota, epoetin omega, epoetin delta, epoetin zeta, epoetin theta, and epoetin delta, PEGylated erythropoietin, carbamylated erythropoietin, and molecules or variants or analogs thereof.

[0147] Among certain exemplary proteins are the specific proteins described below, including fusions, fragments, analogs, variants, or derivatives thereof: OPGL-specific antibodies, peptibodies, and related proteins, etc. (also referred to as RANKL-specific antibodies, peptibodies, etc.), including fully humanized and human OPGL-specific antibodies, particularly fully humanized monoclonal antibodies; myostatin-binding proteins, peptibodies, related proteins, etc., including myostatin-specific peptibodies; IL-4 receptor-specific antibodies, peptibodies, related proteins, etc. Proteins, etc., particularly those that inhibit the activity mediated by binding of IL-4 and / or IL-13 to their receptors; interleukin 1-receptor 1 ("IL1-R1")-specific antibodies, peptibodies, related proteins, etc.; Ang2-specific antibodies, peptibodies, related proteins, etc.; NGF-specific antibodies, peptibodies, related proteins, etc.; CD22-specific antibodies, peptibodies, related proteins, etc., particularly dimers of human-mouse monoclonal hLL2 gamma chain disulfide bound to human-mouse monoclonal hLL2 kappa chain. human CD22-specific antibodies, including but not limited to humanized and fully human monoclonal antibodies, particularly including but not limited to human CD22-specific IgG antibodies, such as the human CD22-specific fully humanized antibody epratuzumab (CAS Registry Number 501423-23-0); IGF-1 receptor-specific antibodies, peptibodies, and related proteins, including but not limited to anti-IGF-1R antibodies; and antibodies that bind to B7RP-1 and its natural receptors on activated T cells. B-7 related protein 1 specific antibodies, peptibodies, related proteins, etc. (also referred to as "B7RP-1", B7H2, ICOSL, B7h, and CD275), including but not limited to, B7RP-specific fully human monoclonal IgG2 antibodies that bind to an epitope in the first immunoglobulin-like domain of B7RP-1, including but not limited to those that inhibit its interaction with its receptor, ICOS; HuMax, e.g., 146B7;IL-15-specific antibodies, peptibodies, related proteins, and the like, including but not limited to IL-15 antibodies and related proteins, particularly humanized monoclonal antibodies; IFN-γ-specific antibodies, peptibodies, related proteins, and the like, including but not limited to human IFN-γ-specific antibodies and fully human anti-IFN-γ antibodies; TALL-1-specific antibodies, peptibodies, related proteins, and the like, as well as other TALL-specific binding proteins; parathyroid hormone ("PTH")-specific antibodies, peptibodies, related proteins, and the like; thrombopoietin receptor ("TPO-R")-specific antibodies, peptibodies, related proteins, etc.; hepatocyte growth factor ("HGF")-specific antibodies, peptibodies, related proteins, etc., including those targeting HGF / SF; cMet axis (HGF / SF:c-Met), such as fully human monoclonal antibodies that neutralize hepatocyte growth factor / scatter (HGF / SF)-specific antibodies; TRAIL-R2-specific antibodies, peptibodies, related proteins, etc.; activin A-specific antibodies, peptibodies, proteins, etc.; TGF-β-specific antibodies, peptibodies, related proteins, etc.; Amyloid beta protein-specific antibodies, peptibodies, related proteins, etc.; c-Kit-specific antibodies, peptibodies, related proteins, etc., including but not limited to proteins that bind to c-Kit and / or other stem cell factor receptors; OX40L-specific antibodies, peptibodies, related proteins, etc., including but not limited to proteins that bind to OX40L and / or other ligands of the OX40 receptor; Activase® (alteplase, tPA), Aimovig® (erenumab), Aranesp® (darbepoetin alfa), Epogen® (epoetin alfa, or erythropoietin), GLP-1, Avonex® (interferon beta-1a), Bexxar® (tositumomab, an anti-CD22 monoclonal antibody), Betaseron® (interferon-beta), Campath® (alemtuzumab, an anti-CD52 monoclonal antibody), Dynepo® (epoetin delta), Velcade® (bortezomib), MLN0002 (anti-alpha4beta7mAb), MLN1202 (anti-CCR2 chemokine receptor mAb), Enbrel® (etanercept, TNF receptor / Fc fusion protein, TNF blocker), Eprex® (epoetin alfa), Erbitux® (cetuximab, anti-EGFR / HER1 / c-ErbB-1), Evenity® (romosozumab), Genotropin® (somatropin, human growth hormone), Herceptin® (trastuzumab, anti-HER2 / neu (erbB2) receptor mAb), Humatrope® (somatropin, human growth hormone), Humira® (adalimumab), Vectibix® (pancreatic ischemia inhibitor), tumumab), Xgeva® (denosumab), Prolia® (denosumab), Enbrel® (etanercept, TNF-receptor / Fc fusion protein, TNF blocker), Nplate® (romiplostim), rilotumumab, ganitumab, conatumumab, brodalumab, insulin in solution, Infergen® (interferon alfacon-1), Natrecor® (nesiritide, recombinant human B-type natriuretic peptide (hBNP), Kineret® (anakinra), Leukine® (sargamostim, rhuGM-CSF), LymphoCide® (epratuzumab, anti-CD22 mAb), Benlysta™ (lymphostat B, belimumab, anti-BlyS mAb), Metalyse® (tenecteplase, t-PA analog), Mircera® (methoxypolyethylene glycol-epoetin beta), Mylotarg® (gemtuzumab ozogamicin), Raptiva® (efalizumab), Cimzia® (certolizumab pegol, CDP870), Soliris™ (eculizumab), pexelizumab (anti-C5 complement), Numax® (MEDI-524), Lucentis® (ranibizumab), Panorex® (17-1A, edrecolomab), Trabio® (lerdelimumab), TheraCimhR3 (nimotuzumab), Omnitarg (pertuzumab, 2C4), Osidem® (IDM-1), OvaRex® (B43.13), Nuvion® (vigilizumab), cantuzumab mertansine (huC242-DM1), NeoRecormon® (epoetin beta), Neumega® (oprelvekin, human interleukin-11), Orthoclone OKT3® (muromonab-CD3, anti-CD3 monoclonal antibody), Procrit® (epoetin alfa), Remicade® (infliximab, anti-TNFα monoclonal antibody), Reopro® (abciximab, anti-GP IL6 receptor monoclonal antibody), Actemra® (anti-IL6 receptor mAb), Avastin® (bevacizumab), HuMax-CD4 (zanolimumab), Rituxan® (rituximab, anti-CD20 mAb), Tarceva® (erlotinib), Roferon-A® (interferon alpha-2a), Simulect® (basiliximab), Prexige® (lumiracoxib), Synagis® (palivizumab), 146B7-CHO (anti-IL15 antibody, see U.S. Pat. No. 7,153,507), Tysabri® (natalizumab, anti-alpha4 integrin mAb), Valortim® (MDX-1303, anti-anthrax protective antigen mAb), ABthrax™, Xolair® (omalizumab), ETI211 (anti-MRSA mAb), IL-1 trap (the Fc portion of human IgG1 and the extracellular domains of both IL-1 receptor components (type I receptor and receptor accessory protein)), VEGF trap (IgG1 Ig domain of VEGFR1 fused to Fc), Zenapax® (daclizumab), Zenapax® (daclizumab, anti-IL-2Rα mAb), Zevalin® (ibritumomab tiuxetan), Zetia® (ezetimibe), Orencia® (atacicept, TACI-Ig), anti-CD80 monoclonal antibody (galiximab), anti-CD23mAb (lumiliximab), BR2-Fc (huBR3 / huFc fusion protein, soluble BAFF antagonist), CNTO148 (golimumab, anti-TNFα mAb), HGS-ETR1 (mapatumumab, human anti-TRAIL receptor-1 mAb), HuMax-CD20 (ocrelizumab, anti-CD20 human mAb), HuMax-EGFR (zalutumumab), M200 (volociximab, anti-α5β1 integrin mAb), MDX-010 (ipilimumab, anti-CTLA-4 mAb, and VEGFR-1 (IMC-18F1), anti-BR3 mAb, anti-C. difficile toxin A and toxin BC mAbs MDX-066 (CDA-1) and MDX-1388), anti-CD22 dsFv-PE38 conjugate (CAT-3888 and CAT-8015), anti-CD25 mAb (HuMax-TAC), anti-CD3 mAb (NI-0401), adecatumumab, anti-CD30 mAb (MDX-060), MDX-1333 (anti-IFNAR), anti-CD38 mAb (HuMax CD38), anti-CD40L mAb, anti-Cripto mAb, anti-CTGF idiopathic pulmonary fibrosis stage 1 fibrogen (FG-3019), anti-CTLA4 mAb, anti-eotaxin 1 mAb (CAT-213), anti-FGF8 mAb, anti-ganglioside GD2 mAb, anti-ganglioside GM2 mAb, anti-GDF-8 human mAb (MYO-029), anti-GM-CSF receptor mAb (CAM-3001), anti-HepC mAb (HuMax HepC), anti-IFNα mAb (MEDI-545, MDX-1103), anti-IGF1R mAb, anti-IGF-1R mAb (HuMax-Inflam), anti-IL12 mAb (ABT-874), anti-IL12 / IL23 mAb (CNTO1275), anti-IL13 mAb (CAT-354), anti-IL2Ra mAb (HuMax-TAC), anti-IL5 receptor mAb, anti-integrin receptor mAb (MDX-018, CNTO95), anti-IP10 ulcerative colitis mAb (MDX-1100), BMS-66513, anti-mannose receptor / hCGβ mAb (MDX-1307), anti-mesothelin dsFv-PE38 conjugate (CAT-5001), anti-PD1 mAb (MDX-1106(ONO-4538)), anti-PDGFRα antibody (IMC-3G3), anti-TGFβmAb(GC-1008), TRAIL-2 mAb (HGS-ETR2), TWEAK mAb, VEGFR / Flt-1 mAb, ZP3 mAb (HuMax-3)

[0148] In some embodiments, the drug delivery device may contain or be used in conjunction with a sclerostin antibody, such as, but not limited to, romosozumab, brosozumab, or BPS 804 (Novartis), or in other embodiments, a monoclonal antibody (IgG) that binds to human proprotein convertase subtilisin / kexin type 9 (PCSK9). Such PCSK9-specific antibodies include, but are not limited to, Repatha® (evolocumab) and Praluent® (alirocumab). In other embodiments, the drug delivery device may contain or be used in conjunction with rilotumumab, bixalomer, trebananib, ganitumab, conatumumab, motesanib diphosphate, brodalumab, bizupiprant, or panitumumab. In some embodiments, the reservoir of the drug delivery device may be loaded with, or the device may be used in conjunction with, IMLYGIC® (talimogene laherparepvec) or another oncolytic HSV for the treatment of melanoma or other cancers, including but not limited to, OncoVEXGALV / CD; OrienX010; G207, 1716; NV1020; NV12023; NV1034; and NV1042. In some embodiments, the drug delivery device may contain or be used in conjunction with, an endogenous tissue inhibitor of metalloproteinase (TIMP), such as, but not limited to, TIMP-3. Antagonistic antibodies of the human calcitonin gene-related peptide (CGRP) receptor, such as, but not limited to, erenumab, and bispecific antibody molecules targeting the CGRP receptor and other headache targets, may also be delivered using the drug delivery devices of the present disclosure. Additionally, bispecific T-cell engager (BiTE®) antibodies, such as, but not limited to, BLINCYTO® (blinatumomab), may be used in or with the drug delivery devices of the present disclosure. In some embodiments, the drug delivery devices may contain or be used with APJ large molecule agonists, such as, but not limited to, apelin or analogs thereof.In some embodiments, a therapeutically effective amount of anti-thymic stromal lymphopoietin (TSLP) or TSLP receptor antibody is used in or in conjunction with the drug delivery devices of the present disclosure.

[0149] Drug delivery devices, assemblies, components, subsystems, and methods have been described in terms of exemplary, but not limited to, embodiments. This detailed description is to be construed as exemplary only and does not describe every possible embodiment of the present disclosure. Various alternative embodiments may be implemented using either current technology or technology developed after the filing date of this patent, and such embodiments will still fall within the scope of the claims that define the invention disclosed herein.

[0150] Those skilled in the art will appreciate that numerous modifications, variations, and combinations can be made to the above-described embodiments without departing from the spirit and scope of the invention disclosed herein, and that such modifications, variations, and combinations are to be construed as being within the scope of the inventive concept.

[0151] Other Considerations While this disclosure provides detailed descriptions of many different embodiments, it should be understood that the legal scope of the description is defined by the language of the claims, and their equivalents, set forth at the end of this patent. This detailed description is intended to be exemplary only and does not describe every possible embodiment, as describing every possible embodiment would be impractical. Various alternative embodiments may be implemented using either current technology or technology developed after the filing date of this patent, and such embodiments would still fall within the scope of the claims.

[0152] The following additional considerations also apply to the above description: Throughout this specification, multiple instances may implement a component, operation, or structure described as a single instance. While individual operations of one or more methods are illustrated and described as separate operations, one or more of the individual operations may be performed simultaneously, and there is no requirement that the operations be performed in the order illustrated. Structures and functions presented as separate components in example configurations may also be implemented as a combined structure or component. Similarly, structures and functions presented as a single component may also be implemented as separate components. These and other modifications, improvements, additions, and refinements are within the scope of the present specification.

[0153] Furthermore, certain embodiments are described herein as including logic circuitry, or a plurality of routines, subroutines, applications, or instructions. These may constitute either software (e.g., code embodied on a machine-readable medium or in a transmission signal) or hardware. Within hardware, routines, etc., are tangible units capable of performing particular operations, and may be configured or arranged in a particular way. In exemplary embodiments, one or more computer systems (e.g., standalone, client, or server computer systems), or one or more hardware modules of a computer system (e.g., a processor or processors), may be configured as hardware modules that operate with software (e.g., an application or application portion) to perform particular operations described herein.

[0154] In various embodiments, a hardware module may be implemented mechanically or electronically. For example, a hardware module may include dedicated circuitry or logic circuitry that is permanently configured to perform specific operations (e.g., as a dedicated processor such as a field programmable gate array (FPGA) or application specific integrated circuit (ASIC)). A hardware module may also include programmable logic or circuitry that is temporarily configured by software to perform specific operations (e.g., as contained in a general-purpose processor or other programmable processor). It will be appreciated that the decision to implement a hardware module mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be determined by cost and time considerations.

[0155] Thus, the term "hardware module" should be understood to encompass a tangible entity, whether physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a particular manner or perform particular operations described herein. When considering embodiments in which hardware modules are temporarily configured (e.g., programmed), each hardware module need not be configured or instantiated at any instance in time. For example, if the hardware modules include a general-purpose processor configured with software, the general-purpose processor may be configured as different hardware modules at different times. Thus, the software may, for example, configure the processor to configure a particular hardware module at one instance and a different hardware module at a different instance.

[0156] As used herein, the term "couple" does not require a direct coupling or connection; two items may be "coupled" to one another through one or more intermediate components or other elements, such as an electronic bus, electrical wiring, mechanical components, or other such indirect connection.

[0157] Hardware modules can provide information to and receive information from other hardware modules. Thus, the described hardware modules can be considered to be communicatively coupled. When multiple such hardware modules are present simultaneously, communication can be achieved by signal transmission (e.g., via appropriate circuits and buses) connecting the hardware modules. In embodiments in which multiple hardware modules are configured or instantiated at different times, communication between such hardware modules can be achieved, for example, through the storage and retrieval of information in a memory structure accessed by the multiple hardware modules. For example, one hardware module can perform an operation and store the output of that operation in a memory device to which it is communicatively coupled. An additional hardware module can subsequently access the memory device at a later time to read and process the stored output. Hardware modules can also initiate communication with input or output devices and operate on resources (e.g., a collection of information).

[0158] Various operations of the example methods described herein may be performed, at least in part, by one or more processors that are temporarily configured (e.g., by software) or permanently configured to perform the operations involved. Whether temporarily or permanently configured, such processors may constitute processor-implemented modules that operate to perform one or more operations or functions. Modules referred to herein may, in some example embodiments, include processor-implemented modules.

[0159] Similarly, the methods or routines described herein may be implemented at least in part by a processor. For example, at least some of the operations of a method may be performed by one or more processors or processor-implemented hardware modules. Performance of certain portions of the operations may be distributed across one or more processors, not just within one machine, but spread across multiple machines. In some exemplary embodiments, one or more processors may be located in a single location, while in other embodiments, the processors may be distributed across multiple locations.

[0160] The performance of certain portions of the operations may be distributed across one or more processors that are located within one machine or that are spread across multiple machines. In some exemplary embodiments, one or more processors or processor-implemented modules may be located in one geographic location (e.g., in a home environment, an office environment, or a server farm). In other embodiments, one or more processors or processor-implemented modules may be distributed across multiple geographic locations.

[0161] This detailed description is to be construed as illustrative only and does not describe every possible embodiment, as describing every possible embodiment would be impractical, if not impossible. Those skilled in the art will be able to implement various alternative embodiments, using either current technology or technology developed after the filing date of this application.

[0162] It will be understood by those skilled in the art that various improvements, modifications and combinations can be made in relation to the above-described embodiments without departing from the scope of the present invention, and such improvements, modifications and combinations are also considered to be included within the scope of the inventive concept.

[0163] The claims at the end of this patent application are not intended to be construed under 35 U.S.C. § 112(f) unless traditional means-plus-function language is explicitly stated, such as "means for" or "step for," unless the claims explicitly state such language. The systems and methods described herein are intended to enhance computer functionality and are intended to improve upon the functionality of conventional computers.< / model> < / model>

Claims

1. 1. A configurable handheld biological analyzer for identifying biological products based on Raman spectroscopy, comprising: a first housing adapted for handheld operation; a first scanner carried by the first housing; a first processor communicatively coupled to the first scanner; and a first computer memory communicatively coupled to the first processor; the first computer memory is configured to load a biological classification model configuration, the biological classification model configuration including a biological classification model, the biological classification model configured to execute on the first processor, the first processor configured to (1) receive a first Raman-based spectral dataset defining a first biological product sample scanned by the first scanner, and (2) use the biological classification model to identify a biological product type based on the first Raman-based spectral dataset; the biological classification model configuration further includes a spectral pre-processing algorithm, and the first processor is configured to execute the spectral pre-processing algorithm when the first Raman-based spectral dataset is received by the first processor to reduce spectral inconsistencies in the first Raman-based spectral dataset; The configurable handheld biological analysis device, wherein the biological classification model includes a classification component selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component.

2. The biological classification model configuration is electronically transferable to a second configurable handheld biological analytical device, the second configurable handheld biological analytical device: a second housing adapted for handheld operation; a second scanner coupled to the second housing; a second processor communicatively coupled to the second scanner; and a second computer memory communicatively coupled to the second processor; the second computer memory is configured to load the biological classification model configuration, the biological classification model configuration including the biological classification model, the biological classification model configured to run on the second processor, the second processor configured to (1) receive a second Raman-based spectral dataset defining a second biological product sample scanned by the second scanner, and (2) use the biological classification model to identify the biological product type based on the second Raman-based spectral dataset; The configurable handheld biological analyzer of claim 1 , wherein the second biological product sample is a new sample of the biological product type.

3. the spectral mismatch is an inter-analytical device spectral mismatch between the first Raman-based spectral dataset and one or more other Raman-based spectral datasets of one or more corresponding other handheld biological analytical devices, each of the one or more other Raman-based spectral datasets representing the biological product type; 10. The configurable handheld biological analyzer of claim 1, wherein the spectral pre-processing algorithm is configured to reduce inter-analytical device spectral discrepancies between the first Raman-based spectral data set and the one or more other Raman-based spectral data sets.

4. The spectral pre-processing algorithm comprises: applying a derivative transform to the first Raman-based spectral dataset to generate a modified Raman-based spectral dataset; aligning the modified Raman-based spectral data set across the Raman shift axis; and normalizing the modified Raman base spectral data set across the Raman intensity axis.

5. 5. The configurable handheld biological analyzer of claim 4, wherein the derivative transformation comprises determining a Raman weighted average of a contiguous group of 11 to 15 Raman shift values across the Raman shift axis and determining a corresponding derivative of these Raman weighted averages across the Raman shift axis.

6. 2. The configurable handheld biological analyzer of claim 1, wherein the classification component is selected to reduce both (1) the Q residual error of the biological classification model and (2) the summary value of the fit of the biological classification model.

7. 10. The configurable handheld biological analyzer of claim 1, wherein the biological classification model further includes a second classification component, and the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component and the second classification component.

8. The configurable handheld biological analyzer of claim 1 , wherein the biological classification model is implemented as a principal component analysis (PCA) model.

9. The configurable handheld biological analyzer of claim 8 , wherein the classification component is a first principal component of the PCA model.

10. The configurable handheld biological analyzer of claim 1 , wherein the computer memory is configured to load a new biological classification model, the new biological classification model including updated classification components.

11. The configurable handheld biological analyzer of claim 1 , wherein the biological classification model configuration is implemented in an Extensible Markup Language (XML) format.

12. The configurable handheld biological analyzer of claim 1 , wherein the biological product type is a therapeutic product.

13. The configurable handheld biological analytical device of claim 1 , wherein the biological product type is identified by the biological classification model during manufacture of a biological product having the biological product type.

14. 10. The configurable handheld biological analyzer of claim 1, wherein the biological classification model is configured to distinguish the first biological product sample having the biological product type from a different biological product sample having a different biological product type based on the classification component.

15. 15. The configurable handheld biological analyzer of claim 14, wherein the biological product type and the different biological product type each have distinct topographical features within the same or similar Raman spectral range.

16. 2. The configurable handheld biological analyzer of claim 1, wherein the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component when the Q residual error or the summary fit value meets a threshold.

17. The configurable handheld biological analyzer of claim 16 , wherein the biological classification model outputs a pass / fail decision based on the threshold value.

18. The configurable handheld biological analyzer of claim 1 , wherein the biological classification model is generated by a remote processor that is remote to the configurable handheld biological analyzer.

19. 1. A bioanalytical method for identifying biological products based on Raman spectroscopy, comprising: loading a biological classification model configuration including the biological classification model into a first computer memory of a first configurable handheld biological analytical device having a first processor and a first scanner; receiving a first Raman-based spectral dataset defining a first biological product sample scanned by the first scanner according to the biological classification model; performing a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the first Raman-based spectral dataset; and identifying a biological product type based on the first Raman-based spectral dataset using the biological classification model; 1. A biological analysis method, wherein the biological classification model includes a classification component selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component.

20. The biological classification model configuration is electronically transferable to a second configurable handheld biological analytical device, and the biological analytical method comprises: loading the biological classification model configuration, including the biological classification model, into a second computer memory of a second configurable handheld biological analytical device having a second processor and a second scanner; receiving a second Raman-based spectral dataset defining a second biological product sample scanned by the second scanner according to the biological classification model; executing the spectral pre-processing algorithm of the biological classification model to reduce a second spectral discrepancy of the second Raman-based spectral dataset; and identifying the biological product type based on the second Raman-based spectral dataset using the biological classification model; 20. The biological analytical method of claim 19, wherein the second biological product sample is a new sample of the biological product type.

21. the spectral mismatch is an inter-analytical device spectral mismatch between the first Raman-based spectral dataset and one or more other Raman-based spectral datasets of one or more corresponding other handheld biological analytical devices, each of the one or more other Raman-based spectral datasets representing the biological product type; 20. The biological analytical method of claim 19, wherein the spectral pre-processing algorithm is configured to reduce the inter-analytical instrument spectral discrepancies between the first Raman-based spectral data set and the one or more other Raman-based spectral data sets.

22. The spectral pre-processing algorithm comprises: applying a derivative transform to the first Raman-based spectral dataset to generate a modified Raman-based spectral dataset; aligning the modified Raman-based spectral data set across the Raman shift axis; and normalizing the modified Raman-based spectral data set across the Raman intensity axis.

23. 23. The biological analytical method of claim 22, wherein the derivative transformation comprises determining Raman weighted averages of consecutive groups of 11 to 15 Raman shift values across the Raman shift axis, and determining corresponding derivatives of these Raman weighted averages across the Raman shift axis.

24. 20. The biological analysis method of claim 19, wherein the classification component is selected to reduce both (1) the Q residual error of the biological classification model and (2) the summary value of the fit of the biological classification model.

25. 20. The biological analytical method of claim 19, wherein the biological classification model further comprises a second classification component, and wherein the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component and the second classification component.

26. 20. The biological analytical method of claim 19, wherein the biological classification model is implemented as a principal component analysis (PCA) model.

27. 27. The biological analytical method of claim 26, wherein the classification component is a first principal component of the PCA model.

28. 20. The biological analysis method of claim 19, wherein the first computer memory is configured to load a new biological classification model, the new biological classification model including an updated classification component.

29. 20. The biological analysis method of claim 19, wherein the biological classification model configuration is implemented in an Extensible Markup Language (XML) format.

30. 20. The biological analytical method of claim 19, wherein the biological product type is a therapeutic product.

31. 20. The biological analytical method of claim 19, wherein the biological product type is identified by the biological classification model during manufacture of a biological product having the biological product type.

32. 20. The biological analytical method of claim 19, wherein the biological classification model is configured to distinguish the first biological product sample having the biological product type from a different biological product sample having a different biological product type based on the classification component.

33. 33. The biological analytical method of claim 32, wherein the biological product type and the different biological product type each have the same or similar Raman spectral range.

34. 20. The biological analytical method of claim 19, wherein the biological classification model is configured to identify the biological product type of the first biological product sample based on the classification component when the Q residual error or the summary fit value meets a threshold.

35. 35. The biological analytical method of claim 34, wherein the biological classification model outputs a pass / fail decision based on the threshold.

36. 20. The biological analytical method of claim 19, wherein the biological classification model is generated by a remote processor that is remote to the configurable handheld biological analytical device.

37. When executed by one or more processors of a configurable handheld biological analyzer, the one or more processors of the configurable handheld biological analyzer: loading a biological classification model configuration into a computer memory of the configurable handheld biological analyzer having a scanner, the biological classification model configuration including a biological classification model; receiving a dataset of Raman-based spectra defining a biological product sample scanned by the scanner according to the biological classification model; performing a spectral pre-processing algorithm of the biological classification model to reduce spectral discrepancies in the Raman-based spectral dataset; and and identifying a biological product type based on the Raman-based spectral dataset using the biological classification model; A tangible, non-transitory computer-readable medium storing instructions for identifying a biological product based on Raman spectroscopy, wherein the biological classification model includes classification components selected to reduce at least one of (1) a Q residual error of the biological classification model, or (2) a summary value of fit of the biological classification model, and the biological classification model is configured to identify the biological product type of the biological product sample based on the classification components.

Images