annotation of target spectroscopy data

Abstract

Embodiments described herein relate to target sample data annotation. A system can include a memory storing computer executable components and a processor executing the computer executable components. The computer executable components can include an encoding component that encodes target molecular data of a target sample into a vectorized format resulting in encoded target molecular data and a matching component that generates a predictive match of the encoded target molecular data to known neutral loss data of a known sample, the known neutral loss data defining mass-to-charge ratio difference values between spectral values of known spectral data corresponding to the known sample.

Description

Background Technology

[0001] Identifying and / or comparing various levels of spectroscopic data from one or more chemical structure measurement devices, data repositories, databases, etc., can be a complex and time-consuming process. One or more variables, such as different data types, data formats, different systems used for data generation, different samples, different execution times and / or lifecycles, and different user subjects, can all affect the ability to accurately and / or efficiently identify and / or compare compounds, fragment ions, and / or neutral loss molecules. In fact, one or more of these variables can lead to problems such as false positives and / or false negatives, low-probability identifications, and a lack of accurate comparisons. In one or more other situations, where manual examination of large amounts of standard spectroscopic data, structural data, molecular data, etc., is required, the execution of identification and / or comparison may be utterly inefficient. Summary of the Invention

[0002] The following summary is presented to provide a basic understanding of one or more exemplary embodiments described herein. This summary is not intended to identify key or essential elements and / or to depict the scope of a particular embodiment or the claims. Its sole purpose is to present concepts in a simplified form as a prelude to a more detailed description that follows. In one or more exemplary embodiments, the systems, computer-implemented methods, apparatuses, and / or computer program products described herein can provide a plug-and-play process for efficiently and automatically annotating unknown or target data using data generated by measuring instruments (also referred to herein as measuring devices) and / or retrieved from data repositories / databases, through structural feature identification, neutral loss identification, fragment ion identification, and / or other spectral features.

[0003] According to one embodiment, a system may include a memory storing computer-executable components and a processor executing the computer-executable components. The computer-executable components may include an encoding component and a matching component; the encoding component encodes target molecular data of a target sample into a vectorized format to obtain encoded target molecular data; the matching component generates a predicted match between the encoded target molecular data and known neutral loss data of a known sample, the known neutral loss data defining the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0004] According to another embodiment, a computer-implemented method may include: encoding target molecular data of a target sample into a vectorized format via a system operatively coupled to a processor to obtain encoded target molecular data; and generating a predictive match between the encoded target molecular data and known neutral loss data of a known sample, the known neutral loss data defining a mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0005] According to another embodiment, a computer program product facilitating a process for annotating target samples may include a computer-readable storage medium embodying program instructions. These program instructions are executable by a processor to encode target molecular data of the target sample into a vectorized format, obtaining encoded target molecular data, and to generate a predicted match between the encoded target molecular data and known neutral loss data of a known sample, wherein the known neutral loss data defines a mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0006] One or more exemplary embodiments described herein may be implemented, associated with, and / or coupled to a chemical structure measurement device, such as a scientific measurement device, like a mass spectrometer.

[0007] One or more exemplary embodiments disclosed herein can be plug-and-play applied to a single measurement device, multiple measurement devices, the same measurement device using multiple replaceable components (e.g., chromatographic columns), etc., to compare output data with unknown data, known data, and / or standard data. The framework described herein can be executed in an efficient and at least partially automated manner, thereby reducing manual processes, improving accuracy, and providing automated reasoning for predictions. In one or more cases, the identification data obtained using one or more exemplary embodiments can be used to construct a database of known molecules, neutral loss, and / or spectral data.

[0008] One or more exemplary embodiments described herein can be used to generate inferred and / or neutral loss data corresponding to a target sample that cannot be obtained simply by comparing the molecular data and / or neutral loss data of the target sample with known molecular data and / or a known neutral loss database. That is, spectral data based on defined spectra may show one or more neutral losses, while another one or more neutral losses may not. In other words, such unmanifested neutral losses may not occur during fragmentation due to chemical structure (e.g., bond type), chemical properties, failure to reach the fragmentation energy, etc. In other words, unmanifested neutral losses may correspond to ions that did not fragment and separate from the target sample for the same and / or different reasons (e.g., chemical structure, chemical properties, fragmentation energy). For example, an unmanifested neutral loss might be a neutral loss that requires a higher energy to be applied to the sample than the energy already applied to break a specific chemical bond.

[0009] One or more exemplary embodiments described herein may be employed to utilize relevant information about one or more compounds that are different from the target compound to be annotated. For example, predictions may be made regarding neutral loss identification (whether it appears or does not appear in the spectral data), predictions regarding ion identification based on chemical structure (e.g., chemical bond type), and / or predictions regarding the identification of the target compound (but not limited to these predictions) based on molecular structure data, neutral loss data, and / or spectral data corresponding to known compounds different from the target compound. For example, such known compounds may belong to the same family, chemical category, etc., as the target compound, and / or may have one or more structural features, ions, and / or neutral losses that are the same as those of the target compound.

[0010] One or more exemplary embodiments described herein can be used to encode data into a universal format that can be used to search, compare, identify, and / or annotate molecular data, spectral data, and / or neutral loss data of a variety of compounds. That is, by using a universal format (such as an encoding format or a vectorized format) discussed below, comparisons can be made in scenarios where such searches, comparisons, identifications, and / or annotations are not possible in existing frameworks. In a non-limiting example, for the same compound, molecular structure data, neutral loss data, and / or spectral data (but not limited to these data) can be encoded, in whole or in part, into the same vectorized format, such as encoded into one or more specific data layers (e.g., including any suitable form of data and / or metadata), thereby enabling efficient comparisons with other such data layers and / or other data (e.g., target compound data) that are also in a vectorized format.

[0011] One or more embodiments described herein may employ one or more such data layers (e.g., including molecular structure data, neutral loss data, and / or spectral data) to compare ion identification results, compound molecular structures, spectral peaks, neutral loss values ​​(e.g., gaps between spectral peaks), etc., of one or more target compounds and / or known compounds. Such comparisons can be used to annotate unknown and / or target compound data and / or generate databases of data layers. Such comparisons can be achieved using databases comprising hundreds, thousands, tens of thousands, or more sets of data layers (but are not limited to this number).

[0012] Furthermore, compared to existing frameworks, this comparison allows for a more comprehensive understanding of the target spectral data. For example, one or more structural and / or neutral loss features can be predicted using a model (such as an artificial intelligence (AI) model or a machine learning (ML) model); this model utilizes the aforementioned database and has learned the correspondences between the molecular structural data, neutral loss data, and / or spectral data included in the database. One or more identified peaks, features, ions, neutral losses, etc., related to known or unknown compounds can be predicted, each corresponding to one or more predicted outputs (such as presented in a sorted and / or weighted format). In one or more cases, the sorted and / or weighted data may be accompanied by one or more sorting and / or weighting rationales based on the correspondences (e.g., correspondences between molecular structural data, neutral loss data, and / or spectral data), and / or these rationales may be provided separately. This helps in understanding the target molecular structural data, neutral loss data, and / or spectral data and their origins, and / or the reasoning behind any one or more identifications provided by the model. Attached Figure Description

[0013] The various embodiments will be readily understood by reading the following detailed description in conjunction with the accompanying drawings. For ease of description, the same reference numerals denote the same structural elements. In the drawings, the embodiments are shown by way of example, not by way of limitation.

[0014] Figure 1 shows a block diagram of an exemplary non-limiting system that facilitates the annotation of target data according to one or more exemplary embodiments described herein.

[0015] Figure 2 shows a block diagram of another exemplary non-limiting system that facilitates the annotation of target data according to one or more exemplary embodiments described herein.

[0016] Figure 3 illustrates a flowchart of the training process of a model that can be executed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0017] Figure 4 illustrates a flowchart of one or more training processes based on neutral loss data, which can be executed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0018] Figure 5 illustrates a flowchart of one or more training processes based on molecular data, which can be executed by the non-limiting system of Figure 2, according to one or more exemplary embodiments described herein.

[0019] Figure 6 illustrates a flowchart of an execution flow that can be performed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0020] Figure 7 shows another flowchart of an execution flow that can be performed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0021] Figure 8 illustrates a flowchart of one or more processes that can be executed by the non-limiting system of Figure 1 according to one or more exemplary embodiments described herein.

[0022] Figure 9 illustrates a flowchart of one or more processes that can be executed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0023] Figure 10 shows a continuation of the flowchart in Figure 9, which involves one or more processes that can be executed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0024] Figure 11 shows another flowchart of one or more processes that can be performed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0025] Figure 12 shows a continuation of the flowchart of Figure 11, which involves one or more processes that can be executed by the non-limiting system of Figure 2 according to one or more exemplary embodiments described herein.

[0026] Figure 13 shows a block diagram of an exemplary operating environment in which embodiments of the topics described herein can be incorporated.

[0027] Figure 14 shows an exemplary schematic block diagram of a computing environment in which the topics described herein can at least partially interact and / or be implemented. Detailed Implementation

[0028] The following detailed descriptions are merely illustrative and are not intended to limit the embodiments and / or their application or utilization. Furthermore, there is no intention to be bound by any express or implied information presented in the foregoing summary or detailed description sections.

[0029] This discussion first focuses on chemical structure measurement equipment, which typically includes, but is not limited to, mass spectrometers and chromatographs. The output of such equipment can be measurement data that defines, but is not limited to, the intensity, mass-to-charge ratio, ionic conductivity, etc., of the analytes, compounds, and / or ions analyzed, eluted, and / or fragmented during the analysis. One type of measurement data is spectroscopic data (also referred to herein as spectral data), generated by the mass spectrometer operation. To compare such spectral data from different analytical batches, multiple compounds, and / or multiple devices, and / or with one or more known and / or standardized datasets, baseline comparisons are more advantageous. This baseline may include using known / control / standard spectroscopic, molecular, and / or neutral loss datasets. However, this approach can be cumbersome, inefficient, and time-consuming when comparing with hundreds or thousands or more analyte standard chromatographic datasets.

[0030] Furthermore, simple comparisons often fail to accurately identify ion fragmentation and / or neutral loss. Going further, existing frameworks of this kind cannot provide one or more predicted peaks, features, ions, neutral loss, etc., nor can they output one or more predictions in a sorted and / or weighted format for each prediction. Moreover, existing frameworks are limited to comparing datasets of the same format, thus preventing such comparisons when datasets are in different formats. Therefore, relatedly, using existing frameworks for dataset annotation may result in the inability to identify ions and / or associated neutral loss arising from fragmentation in the sample, false positives, false negatives, and / or the inability to determine a more accurate output from two or more identification results.

[0031] To overcome one or more of the above-mentioned deficiencies, one or more embodiments described herein may provide a process that utilizes learned correspondences (but not limited to) between molecular structure data, neutral loss data, and / or spectral data to predict one or more molecular structure features, neutral loss, and / or fragment ion identification for a set of chemical data defining a target sample.

[0032] As used herein, molecular structure data can refer to data that defines the chemical structure of a compound, including but not limited to ring structure, chemical bonds, electron pairing, polarity, affinity, charge, hydrophobicity, and hydrophilicity. For example, a specific type of chemical bond associated with a particular charged atom can be a feature of molecular structure data that corresponds to an unmanifested loss of neutrality (e.g., a loss of neutrality that requires a higher energy to be applied to the sample than has been applied).

[0033] Spectral data can refer to data including various value types, such as mass-to-charge ratio (e.g., m / z), conductivity, ionic strength, activation energy, absorbance, etc. For example, the spectrum produced by applying activation energy to a mass spectrometer can be plotted as a graph of ionic strength or absorbance as a function of m / z.

[0034] Neutral loss data can usually be inferred, at least partially, from spectral data. Neutral loss data can refer to the numerical differences between peaks in spectral data. That is, neutral loss can refer to the loss of ions (such as water, hydrogen, etc.) from a compound, which are not represented by m / z peaks but are indicated by the spacing, gaps, and differences between peaks in the spectral data. Neutral loss can appear between adjacent and / or non-adjacent peaks (and / or in specific fragmentation stages or multi-stage mass spectrometry (MS)). n (In the spectrum) peaks (where n values ​​are expected to appear in a specific phase of the spectrum but have not yet been resolved), these peaks may include precursor ions and / or fragment ions. Note that fragment ions may contain one or more elements, atom types, etc.

[0035] Furthermore, in this article, chemical data can refer to any one or more of molecular structure data, neutral loss data, and / or spectral data. For example, the data level described herein refers to data describing a known sample in a format that combines at least the neutral loss data and molecular structure data of that known sample.

[0036] Therefore, such correspondences between molecular structure data, neutral loss data, and / or spectral data can be obtained from known chemical data and databases generated from them, and models can be trained on this database to identify these correspondences. The model can be an artificial intelligence (AI) model, such as a machine learning (ML) model. The AI ​​or ML models used in this paper may include, but are not limited to, any one or more types of models such as neural networks, directed neural networks, convolutional neural networks, image models, and language models.

[0037] In other words, the identification of peaks, neutral loss, etc., can be based on multiple considerations, including but not limited to any one or more different values ​​and / or inferences supported by the molecular structure data, neutral loss data and / or spectral data used by the model.

[0038] Furthermore, as will be described in detail below, the use of a common encoding for molecular structure data, neutral loss data, and / or spectral data facilitates model training and execution. Therefore, the model can evaluate these different types of data to achieve accurate judgments and / or predictions, including multiple comparative outputs that can be sorted, weighted, and / or interpreted based on the correspondences learned and adopted by the model.

[0039] As used herein, unless otherwise stated, the phrase “based on” should be understood to mean “at least partially based on”.

[0040] As used in this article, the term "compound" can refer to a single material, multiple materials, a composition, a sample, a solution, a product, etc.

[0041] As used in this article, the term "data" may include metadata.

[0042] As used herein, the terms “entity,” “requesting entity,” and “user entity” can refer to a machine, device, component, hardware, software, intelligent device, group, organization, individual, and / or human being.

[0043] One or more exemplary embodiments will now be described with reference to the accompanying drawings, wherein the same reference numerals are used to refer to the same drawing elements throughout. In the following description, numerous specific details are set forth for purposes of explanation and are intended to provide a more thorough understanding of the one or more exemplary embodiments. However, it will be apparent that, in various cases, one or more exemplary embodiments may be practiced without these specific details.

[0044] Furthermore, it should be understood that the embodiments shown in one or more of the accompanying drawings described herein are for illustrative purposes only, and therefore the architecture of the embodiments is not limited to the systems, devices and / or components shown therein, nor to any particular order, connection and / or coupling of the systems, devices and / or components shown therein.

[0045] Referring now to Figures 1 and 2, in one or more exemplary embodiments, the non-limiting systems 100 and / or 200 and their systems shown in Figures 1 and 2 may further include one or more computers and / or computing-based elements described herein with reference to the computing environment (computing environment 1400 shown in Figure 14). In one or more of the described embodiments, the computers and / or computing-based elements may be used to implement the operations of one or more systems, devices, components and / or computers shown and / or described in conjunction with Figures 1 and / or 2 and / or other figures herein.

[0046] Turning first to Figure 1, an exemplary, non-limiting system 100 is illustrated, which may include a target data annotation system 102 and a library data repository (DS) 135. In one or more embodiments, known chemical data (known spectral data 132, known neutral loss data 136) corresponding to a known sample 122 may be obtained from the library data repository 135. Optionally, the non-limiting system 100 may include a measurement device (e.g., a chromatography apparatus, mass spectrometer, or other scientific measurement device) from which known chemical data and / or target chemical data (e.g., target molecule data 126) can be obtained. In one or more other embodiments, the measurement device and / or the library data repository 135 may be located external to the target data annotation system 102, which may be communicatively coupled to the measurement device and / or the library data repository 135.

[0047] It should be noted that only a brief description of the target data annotation system 102 is provided as an introductory overview of the more complex and / or more comprehensive target data annotation system 202 shown in FIG. 2. In other words, further details regarding the processes that can be performed by one or more exemplary embodiments described herein will be provided below with respect to the non-limiting system 200 of FIG. 2.

[0048] Referring again to Figure 1, the target data annotation system 102 typically facilitates the analysis of target chemical data, thereby predicting one or more identifications of fragment ions, neutral loss, and / or the target sample itself based on learned correspondences between various types of known chemical data. Such known chemical data types may include, but are not limited to, known molecular data, known spectral data, and / or known neutral loss data.

[0049] As used herein, molecular structure data can refer to data that defines the chemical structure of a compound, including but not limited to ring structure, chemical bonds, electron pairing, polarity, affinity, charge, hydrophobicity, and hydrophilicity. For example, a specific type of chemical bond associated with a particular charged atom can be a feature of molecular structure data that corresponds to an unmanifested loss of neutrality (e.g., a loss of neutrality that requires a higher energy to be applied to the sample than has been applied).

[0050] Spectral data can refer to data including various value types, such as mass-to-charge ratio (e.g., m / z), conductivity, ionic strength, activation energy, absorbance, etc. For example, the spectrum produced by applying activation energy to a mass spectrometer can be plotted as a graph of ionic strength or absorbance as a function of m / z.

[0051] Neutral loss data can usually be inferred, at least partially, from spectral data. Neutral loss data can refer to the numerical differences between peaks in spectral data. That is, neutral loss can refer to the loss of ions (such as water, hydrogen, etc.) from a compound, which are not represented by m / z peaks but are indicated by the spacing, gaps, and differences between peaks in the spectral data. Neutral loss can appear between adjacent and / or non-adjacent peaks (and / or in specific fragmentation stages or multi-stage mass spectrometry (MS)). n (In the spectrum) peaks (where n values ​​are expected to appear in a specific phase of the spectrum but have not yet been resolved), these peaks may include precursor ions and / or fragment ions. Note that fragment ions may contain one or more elements, atom types, etc.

[0052] Furthermore, in this article, chemical data can refer to any one or more of molecular structure data, neutral loss data, and / or spectral data. For example, the data level described herein refers to data describing a known sample in a format that combines at least the neutral loss data and molecular structure data of that known sample.

[0053] The target data annotation system 102 may include at least one memory 104, a bus 105, a processor 106, an encoding component 110, and / or a matching component 120. The processor 106 may be the same as, included in, or different from the processor 1404 (FIG. 14). The memory 104 may be the same as, included in, or different from the system memory 1406 (FIG. 14).

[0054] Using the aforementioned components, the target data annotation system 102 can facilitate one or more comparisons between known chemical data and target chemical data, thereby generating one or more identifications for one or more predicted matches 170. Predictive matches 170 can be for peaks in the spectral data of the target chemical data, neutral losses in the neutral loss data of the target chemical data (corresponding to the spectral data of the target chemical data), and / or the target sample 124 (e.g., the precursor) itself. This can be achieved, for example, regardless of whether the system 102 has been provided with pre-identification information of the target sample 124 and / or input data including any spectral peak from the spectral peaks obtained by analyzing the target sample 124 on a mass spectrometer. This can also be achieved by encoding the target molecular data 126 corresponding to the target sample 124 using a vectorized format 128, which can also be used to encode known chemical data (e.g., known spectral data 132 and / or known neutral loss data 136).

[0055] Typically, the encoding component 110 can encode the target molecular data 126 of the target sample 124 into a vectorized format 128, resulting in encoded target molecular data 130. For example, one or more atoms, ions, chemical bonds, ring structures, number of electrons, charge, polarity, and / or other structural levels can be encoded into the vectorized format 128.

[0056] In one or more embodiments, the vectorization format 128 may include a bit vector generated by the encoding component 110 corresponding to the target molecule data 126. This bit vector may be based on a data fingerprint corresponding to the target molecule data 126 and may include multiple bits representing one or more structural layers of the target sample 124. Figure 3 provides an exemplary fingerprint schematic 301 illustrating the identification of various structural layers corresponding to the target sample 124 in Figure 3.

[0057] In one or more embodiments, the fingerprint data used may be sunlight fingerprint data or other suitable fingerprint data, which is encoded by performing a Tanimoto exponent calculation on the bit vector representation of the structural layers of the target sample 124. That is, the generated bit vector may encode one or more structural layers of the target sample 124 molecule, and optionally, may also encode structural layers of other functional groups coupled / bonded to the target sample 124.

[0058] In one or more cases, the encoding component 110 can be used to verify whether the vectorization format 128 meets one or more requirements, attributes, standards, values, limitations, thresholds, etc., to ensure that the encoded target molecule data 130 can be seamlessly compared with known chemical data (e.g., known spectral data 132 and / or known neutral loss data 136).

[0059] Using the encoded target molecule data 126, the matching component 120 can typically generate a predicted match 170 between the encoded target molecule data 126 and the known neutral missing data 136 of the known sample 122, which defines the mass-to-charge ratio difference 138 between the spectral value 134 of the known spectral data 132 corresponding to the known sample 122.

[0060] Spectral value 134 may include peaks, such as adjacent peaks, non-adjacent peaks, and / or peaks corresponding to fragment ions and / or precursor ions.

[0061] In one or more embodiments, known neutral missing data 136 may be included in known spectral data 132. The known neutral missing data 136 and known spectral data 132 may be provided in any suitable format including data and / or metadata. In one or more embodiments, the known neutral missing data 136 and / or known spectral data 132 may be present at least partially in an encoded format (such as vectorized format 128).

[0062] In other words, by using a common format (such as an encoding format or a vectorized format), comparisons can be made in scenarios where such searches, comparisons, identifications, and / or annotations are not possible within existing frameworks. In a non-limiting example, for the same compound / sample, molecular structure data, neutral loss data, and / or spectral data (but not limited to these) can be encoded, in whole or in part, in the same vectorized format 128, such as encoded in one or more specific data layers (e.g., including any suitable form of data and / or metadata), thereby enabling efficient comparisons with other such data layers and / or other data (e.g., target compound data) that are also in vectorized format 128.

[0063] Predictive match 170 may include neutral loss identification. However, in one or more additional and / or alternative cases, predictive match 170 may additionally and / or alternatively include ion fragment identification, precursor identification, and / or target sample identification. Such identification may be based on inferences and / or correspondences between different types of data. For example, comparing known neutral loss data 136 with target neutral loss data of target sample 124 typically does not yield neutral loss identification as predictive match 170. Instead, peak values, structure values, neutral loss values, and / or other data may be synthesized from known sample 122, which may be the same as or different from target sample 124, and / or includes fragment ions and / or precursors that are the same as and / or different from target sample 124. Therefore, direct comparison may not be accurate, as is done in the existing framework. In contrast, the indirect, inference-based and / or correspondence-based methods employed by the target data annotation system 102 (which will be further described below in conjunction with the target data annotation system 202) can be used to identify one or more predictive matches 170 with higher accuracy and / or interpretability.

[0064] Encoding component 110 and / or matching component 120 are operatively coupled to processor 106, which is operatively coupled to memory 104. Bus 105 enables operative coupling. Processor 106 facilitates the execution of encoding component 110 and / or matching component 120. Encoding component 110 and / or matching component 120 may be stored in memory 104.

[0065] Typically, the unrestricted system 100 can employ any suitable communication method (e.g., electronic, communication, Internet, infrared, fiber optic, etc.) to enable communication between the target data annotation system 102 and any device (such as measuring device 150, like a mass spectrometer) associated with the user entity.

[0066] It should be noted that one or more measuring devices may be communicatively coupled to and / or included in the non-limiting system 100. For example, a first measuring device may have performed spectroscopic analysis on a first compound (a target or known compound), and a second measuring device may have performed spectroscopic analysis on either the first compound or a second compound (another target or known compound). As another example, the first measuring device may have performed spectroscopic analysis on a first target compound (e.g., target sample 124) to obtain target molecule data 125 and associated target spectral data; and the second measuring device may have performed spectroscopic analysis on a second known compound (e.g., known sample 122) to obtain known spectral data 132 and associated known molecule data.

[0067] As a summary of the components and their functions described above, reference is made only briefly to Figure 8, which illustrates a flowchart of an exemplary non-limiting method 800 that facilitates chemical data comparison and target data annotation processes, consistent with one or more exemplary embodiments described herein (e.g., Figure 1 (Non-limiting system 100). Although the non-limiting method 800 is described with respect to the non-limiting system 100 of FIG1, the non-limiting method 800 may also be applied to other systems described herein, such as the non-limiting system 200 of FIG2. For the sake of brevity, repeated descriptions of similar elements and / or processes used in the various embodiments have been omitted.

[0068] At 802, the non-limiting method 800 may include: encoding target molecular data (e.g., target molecular data 126) of a target sample (e.g., target sample 124) into a vectorized format (e.g., vectorized format 128) by a system (e.g., encoding component 110) to obtain encoded target molecular data (e.g., encoded target molecular data 130).

[0069] At 804, the non-limiting method 800 may include: determining, by a system (e.g., encoding component 110 and / or processor 106), whether the vectorization format of the encoded target molecule data has been verified, such as by comparing it with a vectorization format used by known chemical data for comparison with the target chemical data. If yes, the non-limiting method 800 may proceed to step 806. If no, the non-limiting method 800 may return to step 802.

[0070] At 806, the non-limiting method 800 may include: generating a predicted match (e.g., predicted match 170) between encoded target molecule data and known neutral missing data (e.g., known neutral missing data 136) of a known sample (e.g., known sample 122), the known neutral missing data defining a mass-to-charge ratio difference (e.g., m / z difference 138) between the spectral values ​​(e.g., spectral value 134) of the known spectral data (e.g., known spectral data 132) corresponding to the known sample.

[0071] Referring now to FIG2, and simultaneously to FIG6, a non-limiting system 200 is shown, which may include a target data annotation system 202 and a library data repository (DS) 235. For the sake of brevity, repeated descriptions of similar elements and / or processes used in the various embodiments are omitted. The description of the embodiment of FIG1 is applicable to the embodiment of FIG2. Similarly, the description of the embodiment of FIG2 is applicable to the embodiment of FIG1.

[0072] In one or more embodiments, the library data repository 235 may be separate from the non-limiting system 200, but may be communicatively coupled to it.

[0073] It should be noted that one or more measuring devices may be communicatively coupled to and / or included in the non-limiting system 200. For example, a first measuring device may have performed spectroscopic analysis on a first compound (a target or known compound), and a second measuring device may have performed spectroscopic analysis on either the first or second compound (another target or known compound). As another example, the first measuring device may have performed spectroscopic analysis on a first target compound (e.g., target sample 602) to obtain target molecule data 610 and associated target spectral data 607; and the second measuring device may have performed spectroscopic analysis on a second known compound (e.g., known sample 302) to obtain known spectral data 306 and associated known molecule data 310.

[0074] Typically, the target data annotation system 202 facilitates the analysis of target chemical data (e.g., target sample data 246 including target molecular data 610, target spectral data 607, and / or target neutral loss data 608) by predicting one or more identifications of fragment ions, neutral loss, and / or the target sample 602 itself based on learned correspondences between various types of known chemical data (e.g., known sample data 236 including known spectral data 306, known neutral loss data 308, and / or known molecular data 310). That is, such known chemical data types may include, but are not limited to, known molecular data, known spectral data, and / or known neutral loss data.

[0075] As used herein, molecular structure data can refer to data that defines the chemical structure of a compound, including but not limited to ring structure, chemical bonds, electron pairing, polarity, affinity, charge, hydrophobicity, and hydrophilicity. For example, a specific type of chemical bond associated with a particular charged atom can be a feature of molecular structure data that corresponds to an unmanifested loss of neutrality (e.g., a loss of neutrality that requires a higher energy to be applied to the sample than has been applied).

[0076] Spectral data can refer to data including various value types, such as mass-to-charge ratio (e.g., m / z), conductivity, ionic strength, activation energy, absorbance, etc. For example, the spectrum produced by applying activation energy to a mass spectrometer can be plotted as a graph of ionic strength or absorbance as a function of m / z.

[0077] Neutral loss data can usually be inferred, at least partially, from spectral data. Neutral loss data can refer to the numerical differences between peaks in spectral data. That is, neutral loss can refer to the loss of ions (such as water, hydrogen, etc.) from a compound, which are not represented by m / z peaks but are indicated by the spacing, gaps, and differences between peaks in the spectral data. Neutral loss can appear between adjacent and / or non-adjacent peaks (and / or in specific fragmentation stages or multi-stage mass spectrometry (MS)). n (In the spectrum) peaks (where n values ​​are expected to appear in a specific phase of the spectrum but have not yet been resolved), these peaks may include precursor ions and / or fragment ions. It should be noted that fragment ions may contain one or more elements, atom types, etc.

[0078] Furthermore, in this article, chemical data can refer to any one or more of molecular structure data, neutral loss data, and / or spectral data. For example, the data level described herein refers to data describing a known sample in a format that combines at least the neutral loss data and molecular structure data of that known sample.

[0079] One or more communications between one or more components of the unrestricted system 200 may be provided via wired and / or wireless means, including but not limited to cellular networks, wide area networks (WANs) (e.g., the Internet) and / or local area networks (LANs). Suitable wired or wireless technologies used to support communications may include, but are not limited to, Wi-Fi, GSM, UMTS, WiMAX, Enhanced General Packet Radio Service (Enhanced GPRS), 3GPP Long Term Evolution (LTE), 3GPP2 Ultra Mobile Broadband (UMB), High Speed ​​Packet Access (HSPA), Zigbee and other 802.XX wireless technologies and / or traditional telecommunications technologies, BLUETOOTH®, Session Initiation Protocol (SIP), ZIGBEE®, RF4CE, WirelessHART, 6LoWPAN (IPv6 over Low Power Radio Area Networks), Z-Wave, Advanced and / or Adaptive Networking Technology (ANT), Ultra Wideband (UWB) standard protocols and / or other proprietary and / or non-proprietary communication protocols.

[0080] The target data annotation system 202 can be associated with a cloud computing environment (such as cloud computing environment 1300 in Figure 13), and can be accessed via that cloud computing environment.

[0081] The target data annotation system 202 may include multiple components. These components may include a memory 204, a processor 206, a bus 205, an encoding component 210, a generation component 212, a comparison component 214, a sorting component 216, a weighting component 218, a matching component 220, a model 222, a notification component 224, and / or a training component 226. Using these components, the target data annotation system 202 can facilitate the execution of a process to generate one or more predictive matches 270 regarding one or more fragment ions, neutral loss, and / or target samples. In one or more embodiments, the target data annotation system 202 may provide one or more such predictive matches 270 in a sorted and / or weighted format. In one or more cases, the sorted and / or weighted data may be accompanied by one or more sorting and / or weighting rationales based on correspondences (e.g., correspondences between molecular structure data, neutral loss data, and / or spectral data), and / or these rationales may be provided separately. This helps in understanding the target molecule structural data, neutral loss data and / or spectral data and their origins, and / or the reasoning behind any one or more of the identifications provided by the model.

[0082] The processor 206, memory 204, and bus 205 of the target data annotation system 202 are discussed below. For example, in one or more exemplary embodiments, the target data annotation system 202 may include the processor 206 (e.g., a computer processing unit, microprocessor, classical processor, quantum processor, and / or similar processor). In one or more exemplary embodiments, components associated with the target data annotation system 202 as described herein with or without reference to one or more accompanying drawings of one or more exemplary embodiments may include one or more computer-readable, writable, and / or executable and / or machine-readable, writable, and / or executable components and / or instructions that can be executed by the processor 206 to provide the performance of one or more processes defined by such components and / or instructions. In one or more exemplary embodiments, the processor 206 may include an encoding component 210, a generation component 212, a comparison component 214, a sorting component 216, a weighting component 218, a matching component 220, a model 222, a notification component 224, and / or a training component 226.

[0083] In one or more exemplary embodiments, the target data annotation system 202 may include a computer-readable storage 204 operatively connected to a processor 206. The storage 204 may store computer-executable instructions that, when executed by the processor 206, cause the processor 206 and / or one or more other components of the target data annotation system 202 (e.g., encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226) to perform one or more operations. In one or more exemplary embodiments, the storage 204 may store computer-executable components (e.g., encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226).

[0084] The target data annotation system 202 and / or its components described herein may be coupled to each other in a communicative, electrical, operational, optical, and / or other manner via bus 205. Bus 205 may include one or more of the following: memory bus, memory controller, peripheral bus, external bus, local bus, quantum bus, and / or other types of buses that may employ one or more bus architectures. One or more of these examples of bus 205 may be employed.

[0085] In one or more exemplary embodiments, the target data annotation system 202 may be coupled (e.g., in a communicative, electrical, operational, optical, and / or similar manner) to one or more external systems (e.g., an electrical output generation system not shown, one or more output targets and / or output target controllers), sources, and / or devices (e.g., classical and / or quantum computing devices, communication devices, and / or similar devices), such as via a network. In one or more exemplary embodiments, one or more components of the target data annotation system 202 and / or the unrestricted system 200 may reside in the cloud and / or may reside locally in a local computing environment (e.g., at a designated location).

[0086] In addition to the processor 206 and / or memory 204 described above, the target data annotation system 202 may also include one or more computer-readable, writable and / or executable and / or machine-readable, writable and / or executable components and / or instructions, wherein when executed by the processor 206, the components and / or instructions can provide the performance of one or more operations specified by such components and / or instructions.

[0087] The other components of the target data annotation system 202 (e.g., encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226) will be discussed next. As mentioned above, in general, the target data annotation system 202 can facilitate a set of procedures for identifying one or more neutral missing, fragmented ions, and / or target samples.

[0088] These processes can be broken down into a set of sub-processes, including but not limited to training model 222 using known sample data 236, using model 222 to perform a comparison between target sample data 246 and known sample data 236, and similarly using model 222 to generate predictive matches 270 and corresponding output data.

[0089] First, it should be noted that in one or more exemplary embodiments, the encoding component 210, the generation component 212, the comparison component 214, the sorting component 216, the weighting component 218, the matching component 220, the model 222, the notification component 224, and / or the training component 226 can be implemented independently without relying on one or more other components among the encoding component 210, the generation component 212, the comparison component 214, the sorting component 216, the weighting component 218, the matching component 220, the model 222, the notification component 224, and / or the training component 226. Additionally and / or alternatively, the encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226 may be included in the advanced analysis component 203; one or more of the following functions of the encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226 may be performed by the advanced analysis component 203; and / or the encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224, and / or training component 226 may be omitted and performed by the advanced analysis component 203. Perform one or more of the following functions of one or more of the omitted encoding component 210, generation component 212, comparison component 214, sorting component 216, weighting component 218, matching component 220, model 222, notification component 224 and / or training component 226.

[0090] As described above, the first set of one or more processes may include training model 222 using known sample data 236. Therefore, referring to FIG3 and continuing in conjunction with FIG2, one or more data layers 340 may be generated, which include data / metadata in at least part of a vectorized format 326 for use by training component 226 to train model 222.

[0091] Data layer 240 may include any suitable amount of data, including known molecular data 310, encoded known molecular data 328, known spectral data 306, and / or known neutral loss data 308. In this document, known neutral loss data 308 may be included in known spectral data 306.

[0092] Encoding component 210 can retrieve known molecular data 310 from library data repository 235, standard databases, customer databases, and / or any suitable output of the mass spectrometer. This known molecular data 310 can describe various known structural levels 304 of a known sample 302. One or more known structural levels 304 may include, but are not limited to, descriptions and / or definitions of ring structures, chemical bonds, electrons, charges, polarity, molecules, atoms, etc.

[0093] The encoding component 210 can analyze the original known molecular data 310 and encode the known molecular data 310 into a vectorized format 326 to obtain the encoded known molecular data 328.

[0094] In one or more embodiments, the vectorization format 326 may include a bit vector 322 generated by the encoding component 210 based on known fingerprint data 320 corresponding to known molecular data 310. The bit vector 322 may be based on the data fingerprint 320 and may include a plurality of bits 324 representing one or more structural layers 304 of the known sample 302. Figure 3 provides an exemplary fingerprint schematic 301 illustrating the identification of various structural layers of the corresponding sample.

[0095] In one or more embodiments, the fingerprint data 320 used may be sunlight fingerprint data or other suitable fingerprint data, which is encoded by performing a Tannimoto index calculation on the bit vector representation of the structural layer 304 of the known sample 302. That is, a bit vector 322 may be generated to encode one or more structural layers 304 of the molecule of the known sample 302, and optionally, structural layers of other functional groups coupled / bonded to the known sample 302 may also be encoded.

[0096] In one or more cases, the encoding component 210 can be used to verify whether the vectorized format 326 meets one or more requirements, attributes, standards, values, limitations, thresholds, etc., to ensure that the encoded known molecular data 328 can be seamlessly integrated and / or compared with other known chemical data (e.g., chemical data of one or more other known samples 302).

[0097] In one or more cases, encoding component 210 may also acquire known spectral data 306 and / or known neutral loss data 308. Known spectral data 306 may include peak values ​​(e.g., m / z values) of fragment ions and / or precursor ions corresponding to each intensity value. The gaps between peaks (whether adjacent or not) may (but are not always) represent neutral losses of molecules and / or atoms lost from known sample 302 when fragment energies are applied to known sample 302 by the mass spectrometer. For example, in one or more cases, a neutral loss of 18 m / z may represent H₂O or water loss from known sample 302.

[0098] It should be noted that the specific neutral loss at a specific location in known spectral data 306 (representing a specific fragment sequence in known sample 302) can be used to define one or more inferences, such as the identification of other neutral losses in known spectral data 306 and / or the identification of precursor / fragment ions.

[0099] In one or more cases, the encoding component 210 may acquire neutral loss data 308 including one or more m / z differences 316 corresponding to one or more neutral losses 318.

[0100] In one or more other cases, generating component 212 may generate such neutral loss data 308 based on known spectral data 306, including one or more m / z differences 316 corresponding to one or more neutral losses 318. That is, a neutral loss 318 may be defined by an m / z difference 316 between a pair of spectral peaks 312, as shown in spectrum 314 of known spectral data 306. This pair of spectral peaks 312 may correspond to fragment ions and / or precursor ions that are adjacent and / or non-adjacent to each other.

[0101] In one or more cases, generation component 212 may generate neutral loss data 308 that is not apparent in the spectrum 314 defined by known spectral data 306. That is, such unseen neutral losses 318 may not have occurred during fragmentation due to chemical structure (e.g., bond type), chemical properties, failure to reach fragmentation energy, etc. In other words, unseen neutral losses may correspond to ions that did not fragment and separate from the target sample for the same and / or different reasons (e.g., chemical structure, chemical properties, fragmentation energy). During this training phase, this data is not inferred but is directly available, for example, as part of known neutral loss data 308 and / or known spectral data 306, and / or provided as input by a user entity using a computing device communicatively coupled to the target data annotation system 202.

[0102] In one or more cases, it is known that neutral missing data 308 may be in an uncoded state and therefore not in vectorized format 326.

[0103] Based on the input and / or generation of known neutral missing data 308 for known sample 302, and the encoding process of known molecular data 328, generation component 212 can generate tag data 342 that associates the known neutral missing data 308 with the encoded known molecular data 328. That is, neutral missing data 318 can be matched with structural level 204 via tags, links, markers, tables, matrices, nodes and edges, and / or any other tag data 342. Thus, the known neutral missing data 308 can be at least partially vectorized, whether provided directly in vectorized format 326 or referenced by tag data 342. In one or more cases, encoding component 210 can assist generation component 212 by encoding one or more levels of neutral missing data 308 into bit vectors 322 (and thus into vectorized format 326).

[0104] Finally, generation component 212 can generate one or more data layers 340, summarizing encoded known molecular data 328, known neutral loss data 308 (whether or not in vectorized format 326), and label data 342 corresponding to known samples. Generation component 212 can store this data layer 340 in a library data repository 235, or in any other suitable location communicatively coupled and / or accessible to the target data annotation system 202.

[0105] The training component 226, along with Figures 4 and 5, will be discussed next. In brief, the training component 226 can train the model 222 using multiple data layers 340 of different known samples 302 (such as data layers 340 divided into known groups, validation groups, and / or training groups). Relatedly, as described below, one or more additional layers of data can be generated, such as labeled data and / or additional neutral loss data 308 corresponding to one or more unmanifested neutral loss 318.

[0106] Model 222 can be an artificial intelligence (AI) model, such as a machine learning (ML) model. The AI ​​model or ML model 322 used in this paper can include any one or more types of models, including but not limited to neural networks, directed neural networks, convolutional neural networks, image models, language models, etc.

[0107] Therefore, typically, training component 226 can use a set of data layers 304, including encoded known molecular data 328, corresponding known neutral loss data 308, and corresponding label data 342, to train one or more models 222.

[0108] For example, referring first to Figure 4, model 222 can be trained to receive input neutral loss data and match it with encoded molecular structure data. Figure 4 shows spectrum 314, which includes spectral values ​​312 and known spectral data 306 based on known sample 302. As shown, in the known spectral data 306, known neutral loss data 308 is exemplified as one or more neutral losses 318. For example, such neutral losses 318 may be explicit and detectable by generation component 212. Figure 4 also shows known molecular data 310 represented by known data fingerprint 320 of known sample 302.

[0109] Based on the data input to model 222 from training component 226, model 222 may, under the guidance of training component 226, generate a set of input neurons 452 and hidden neurons 454 corresponding to known neutral missing data 308, and an additional set of input neurons 458 corresponding to encoded known molecular structure data 310. In response, model 222 may generate a set of one or more output neurons 456 containing the summarized encoded known molecular structure data 310 and neutral missing data 308. Output neurons 456 may be labeled to one or more data layers 340 of the known samples 302, and / or included in these data layers.

[0110] Referring next to Figure 5, model 222 can also be trained to obtain the input encoded molecular structure data and match the input encoded molecular structure data with neutral loss data. Figure 5 shows the known data fingerprint 320 of the known sample 320, which represents the known molecular data 310. Figure 5 also shows the known spectral data 306 and / or the known neutral loss data 308, which are represented as one or more neutral losses 318 in the spectrum 314.

[0111] Based on the data input to model 222 from training component 226, model 222, under the guidance of training component 226, can generate a set of input neurons 552 and hidden neurons 554 corresponding to the encoded known molecular structure data 310, and a set of additional input neurons 558 corresponding to the known neutral loss data 308. In response, model 222 can generate a set of one or more output neurons 556, which includes the summarized encoded known molecular structure data 310 and neutral loss data 308. Output neurons 556 can be labeled to one or more data layers 340 of the known samples 302, and / or included in these data layers.

[0112] It should be noted that although Figures 4 and 5 involve model 222, which includes neural networks and / or is itself a neural network, model 222 may also employ and / or include one or more other types of models, such as other related models. Therefore, the illustrations and descriptions of Figures 4 and 5 are not limiting and are more broadly used to illustrate how model 222 summarizes different types of data during the training process with the assistance of training component 226.

[0113] It should also be noted that, as previously described and illustrated in the neuron diagrams of Figures 4 and 5, in one or more cases, model 222 may learn one or more additional learned neutral losses 318 (such as unmanifested neutral losses 318) based on the generation of one or more data layers 340. That is, based on the aggregated data that associates molecular structural features with neutral losses in at least a partially vectorized format 326, one or more additional neutral losses 318 may be learned through data overlap, intrinsic correlation, matching correspondence, etc.

[0114] Additionally and / or alternatively, in one or more embodiments, the training component 226 may facilitate feedback evaluation of prediction matches 270 relative to one or more outputs. For example, this may include a user entity (e.g., using a computing device communicatively coupled to the non-limiting system 200) inputting data to request or modify one or more weights of one or more model hyperparameters of one or more trained models 222.

[0115] The second set of processes is discussed next, which is used to execute the trained model 222 to predict one or more prediction matches 270.

[0116] That is, referring to Figure 6 and continuing in conjunction with Figure 2, it is expected that the trained model 222 will generate one or more predictive matches 270 related to one or more target samples 602. For example, the expected identification may include target sample identification, precursor identification, fragment peak identification, and / or neutral loss identification, any of which may be included in one or more predictive matches 270 to be generated by the trained model 222 in conjunction with the target data annotation system 202.

[0117] Encoding component 210 can acquire target molecule data 610 from library data repository 235, standard databases, customer databases, any suitable output of a mass spectrometer, and / or any other computer device associated with a user entity and communicatively coupled to non-limiting system 200. This target molecule data 610 can describe various target structural levels 604 of the target sample 602. One or more target structural levels 604 may include, but are not limited to, descriptions and / or definitions of rings, chemical bonds, electrons, charges, polarity, molecules, atoms, etc.

[0118] The encoding component 210 can analyze the original target molecule data 610 and encode the target molecule data 610 into a vectorized format 326 to obtain the encoded target molecule data 628.

[0119] In one or more embodiments, the vectorization format 326 (e.g., the same format used by the non-limiting system 200 described above for known sample data 236) may include a bit vector 622 generated by the encoding component 210 based on target fingerprint data 620 corresponding to the target molecule data 610. The bit vector 622 may be based on the data fingerprint 620 and may include a plurality of bits 624 representing one or more structural layers 604 of the target sample 602. Figure 3 provides an exemplary fingerprint schematic 301 illustrating the identification of various structural layers of the corresponding sample.

[0120] In one or more embodiments, the fingerprint data 620 used may be sunlight fingerprint data or other suitable fingerprint data, which is encoded by performing a Tanimoto exponent calculation on the bit vector representation of the structural layer 604 of the target sample 602. That is, a bit vector 622 may be generated that encodes one or more structural layers 604 of the molecule of the target sample 602, and optionally, structural layers of other functional groups coupled / bonded to the target sample 602 may also be encoded.

[0121] In one or more cases, the encoding component 210 can be used to verify whether the vectorization format 326 meets one or more requirements, attributes, standards, values, limits, thresholds, etc., to ensure that the encoded target molecular data 628 can be seamlessly integrated and / or compared with other target chemical data (e.g., target chemical data of one or more other target samples 602 and / or known sample 302).

[0122] Thus, data can be encoded in a universal format using one or more exemplary embodiments described herein, which can be used for searching, comparing, identifying, and / or annotating molecular data, spectral data, and / or neutral loss data related to various compounds. That is, by using universal formats (such as encoding formats or vectorized formats) discussed below, comparisons can be made in scenarios where such searching, comparing, identifying, and / or annotating is not possible in existing frameworks. In a non-limiting example, for the same compound, molecular structure data, neutral loss data, and / or spectral data (but not limited to these data) can be encoded, in whole or in part, in the same vectorized format, such as encoded in one or more specific data layers (e.g., including any suitable form of data and / or metadata), thereby enabling efficient comparison with other such data layers and / or other data (e.g., target compound data) that are also in a vectorized format.

[0123] In one or more cases, encoding component 210 may also acquire target spectral data 606 and / or target neutral loss data 608. Target spectral data 606 may include peak values ​​(e.g., m / z values) of fragment ions and / or precursor ions corresponding to each intensity value. The gaps between peaks (whether adjacent or not) may (but are not always) represent neutral losses corresponding to molecules and / or atoms lost by the target sample 602 when fragment energies are applied to the target sample 602 by the mass spectrometer. For example, in one or more cases, a neutral loss of 18 m / z may represent the loss of H2O (or water) in the target sample 602.

[0124] It should be noted that the specific neutral loss at a specific location in the target spectral data 606 (representing a specific fragment sequence of the target sample 602) can be used to define one or more inferences, such as the identification of other neutral losses in the target spectral data 606 and / or the identification of precursor / fragment ions.

[0125] In one or more cases, the encoding component 210 may acquire neutral loss data 608 containing one or more mass-to-charge ratio differences 616 corresponding to one or more neutral losses 618.

[0126] In one or more other cases, generating component 212 may generate such neutral loss data 608 based on target spectral data 606, including one or more m / z differences 616 corresponding to one or more neutral losses 618. That is, a neutral loss 618 may be defined by an m / z difference 616 between a pair of spectral peaks 612, as shown in spectrum 614 of target spectral data 606. This pair of spectral peaks 612 may correspond to fragment ions and / or precursors that are adjacent and / or non-adjacent to each other.

[0127] In one or more cases, generation component 212 may generate target neutral loss data 608 that is not apparent in the spectrum 614 defined by target spectral data 606. That is, such unseen neutral losses 618 may not occur during fragmentation due to chemical structure (e.g., bond type), chemical properties, failure to reach fragmentation energy, etc. In other words, unseen neutral losses may correspond to ions that did not fragment and separate from the target sample for the same and / or different reasons (e.g., chemical structure, chemical properties, fragmentation energy). During this training phase, this data is not inferred but is directly available, for example, as part of the target neutral loss data 608 and / or target spectral data 606, and / or provided as input by a user entity using a computing device communicatively coupled to the target data annotation system 202.

[0128] Using target sample data 246 (e.g., target molecule data 610, encoded target molecule data 628, target neutral loss data 608, and / or target spectral data 607) as the first input, and using the known sample data 236 discussed above, the comparison component 214, the sorting component 216, the weighting component 218, and / or the matching component 220 can perform one or more processes.

[0129] In one or more cases, one or more of the comparison component 214, sorting component 216, weighting component 218, and / or matching component 220 may be included in model 222. In one or more other cases, one or more processes performed by one or more of the comparison component 214, sorting component 216, weighting component 218, and / or matching component 220, as described below, may be performed by model 222. In one or more cases, one or more of the comparison component 214, sorting component 216, weighting component 218, and / or matching component 220 may be omitted, thereby integrating the executed processes into the functionality of model 222. In one or more cases, one or more of the comparison component 214, sorting component 216, weighting component 218, and / or matching component 220 may be non-physical components representing one or more functions of model 222.

[0130] First, consider comparison component 214, which typically compares known sample data 236 with target sample data 246. In one or more embodiments, comparison component 214 can compare data of the same type (e.g., encoded known molecular data 328 with encoded target molecular data 628) and / or data of different types (e.g., known neutral loss data 308 with encoded target molecular data 628). For the latter example, a neutral loss 318 in the known neutral loss data 308 can be matched and / or compared with a structural feature in the encoded target molecular data 628 corresponding to the ion loss involved in that neutral loss 318 (e.g., target structural layer 604).

[0131] For example, one or more exemplary embodiments described herein may be employed to utilize relevant information from one or more compounds that are different from the target compound to be annotated. That is, predictions regarding neutral loss identification (whether it appears or does not appear in the spectral data), ion identification based on chemical structure (e.g., bond type), and / or target compound identification (but not limited to these predictions) may be made based on molecular structure data, neutral loss data, and / or spectral data (but not limited to these data) corresponding to known compounds different from the target compound. For example, such known compounds may belong to the same family, chemical category, etc., as the target compound, and / or may have one or more structural features, ions, and / or neutral losses that are the same as the target compound.

[0132] In one or more cases, the comparison may include comparing the aggregated known sample data 236 (e.g., data included in data layer 340) with any one or more types of target sample data 246. That is, this is a more significant advantage provided by one or more embodiments described herein compared to existing frameworks.

[0133] For example, one or more exemplary embodiments described herein can be used to generate inference and / or neutral loss data 608 corresponding to target sample 602 via the above comparison, whereas these inference and / or neutral loss data cannot be obtained by simply comparing the molecular data and / or neutral loss data of target sample 602 with a known molecular data 310 library (e.g., non-coded data) and / or a known neutral loss data 308 library (e.g., non-coded data).

[0134] Another example is that, based on spectral data defining the spectrum (whether it is target data and / or known data), one or more neutral losses may be present, while one or more other neutral losses may not be present. That is, such unpresented neutral losses may not occur during fragmentation due to chemical structure (e.g., bond type), chemical properties, failure to reach fragmentation energy, etc. In other words, unpresented neutral losses may correspond to ions that did not fragment and separate from the target sample for the same and / or different reasons (e.g., chemical structure, chemical properties, fragmentation energy). Such unpresented neutral losses corresponding to target sample 602 can be determined by using aggregated known sample data 236 (such as aggregated known sample data in one or more data layers 340 corresponding to one or more known samples 302, which may be the same as and / or different from target sample 602).

[0135] For example, although a known sample 302 that is different from the target sample 602 may contain different ions, structural layers and / or molecules, inferences can be made based on similarity, such as due to one or more structural layers 304, 604 (e.g., chemical bond type and / or position, etc.), to infer the unmanifested neutral loss that is predicted to correspond to the target sample 602.

[0136] Based on the comparisons provided by comparison component 214, matching component 220 can typically generate a predicted match 270 between at least one level of known sample data 236 and target sample data 246. For example, matching component 220 can generate a predicted match 270 based on encoded target molecular data 628 and known neutral loss data 308 of known sample 302, where known neutral loss data 308 defines a mass-to-charge ratio difference 316 between the spectral values ​​312 of known spectral data 306 corresponding to known sample 302. That is, consistent with what has been discussed directly above, this predicted match 270 can be based on the use of encoded summary data of known sample 302 (e.g., encoded summary data in one or more data levels 340).

[0137] Another example is that the matching component 220 can match known neutral missing data 308 with known bits 624 in the target bit vector 622.

[0138] As previously described, the predictive match 270 may include one or more identifications, such as the identification of neutral losses corresponding to the target sample 602 (such as unmanifested neutral losses, e.g., neutral losses not manifested in the known spectrum 614). Additionally and / or alternatively, the predictive match 270 may include the identification of the target sample 602, the precursor corresponding to the target sample 602, and / or the fragment ions corresponding to the target sample 602.

[0139] In one or more specific cases, predictive matching 270 may include neutral loss identification. However, in one or more additional and / or alternative cases, predictive matching 270 may additionally and / or alternatively include fragment ion identification, precursor identification, and / or target sample identification.

[0140] To reiterate, any one or more of these identifications can be based on inferences and / or correspondences between different types of known sample data 236. For example, comparing known neutral loss data 308 with target neutral loss data of target sample 602 typically does not yield a neutral loss identification as a predicted match 270. Instead, peak values, structure values, neutral loss values, and / or other data can be aggregated from known samples 302 that are the same as or different from target sample 602 and / or include the same and / or different fragment ions and / or precursors compared to target sample 602. Therefore, direct comparisons may not be accurate, as is done in existing frameworks. In contrast, the indirect, inference-based, and / or correspondence-based approach employed by target data annotation system 202 can be used to identify one or more predicted matches 270 with associated higher accuracy and / or interpretability.

[0141] Turning now to Figure 7, and continuing to refer to Figures 2 and 6, in conjunction with comparison component 214, one or more predictive matches 270 may be generated by using sorting 704, reordering 708, and / or weighting 714.

[0142] The ranking 704 and / or re-ranking 708 used in this paper can refer to the ranking of identified annotated samples based on the similarity level with known fragment ions, known neutral loss, and / or known samples corresponding to target fragment ions, target neutral loss, and / or target samples. It should be noted that the ranking is not necessarily a ranking of the same type against the same type (e.g., sample against sample). Rather, for example, a similarity ranking of neutral loss and samples can represent an inference of similarity between the two.

[0143] In contrast, weight 714 may refer to a quantitative annotation of the predictive accuracy of match 270 based on known sample data 236 used to generate a summary of predicted matches 270, such as through matching component 220 and / or trained model 222.

[0144] Accordingly, for example, the ranking component 216 may generate a ranking 704 for a set of one or more known samples 302 based on a first similarity level 702 between the encoded known molecular data 328 corresponding to one or more known samples 302 and the encoded target molecular data 628. Accordingly, the ranking 704 may define a similarity order based on the compared encoded molecular data 328, 628 (e.g., the highest ranking indicates the highest similarity). The ranking 704 may refer to the similarity of a specific identification, and / or be based on the similarity between the known sample 302 and the target sample 602 for each identification.

[0145] In the case of such sorting, comparison component 214 can compare known neutral missing data 308 with target neutral missing data 608 of target sample 602 to obtain a set of one or more possible matches (e.g., possible matches in a set of matches 630), and matching component 220 can generate one or more predicted matches 270 between known sample 302 (e.g., its data) and target sample 602.

[0146] Relatedly, the sorting component 216 may further generate one or more reorderings 708 for the group of one or more known samples 302 (e.g., their data) based on a second similarity level 706 between known neutral missing data 308 corresponding to the group of one or more known samples 302 and target neutral missing data 608 corresponding to the target sample 602. For example, the second similarity level 706 may be applied to neutral missing data in the known neutral missing data 308 that corresponds to a known bit 324 of the encoded known molecular data 328, which matches a target bit 624 of the encoded target molecular data 628. Accordingly, the reordering 708 may define a similarity order (e.g., the highest order indicates the highest similarity) based on a specific set of matches 630 obtained from the initial sorting 704. The reordering 708 may refer to the similarity of a specific identification and / or the similarity between the known sample 302 and the target sample 602 based on each identification.

[0147] It should be noted that in one or more embodiments, reordering can be performed without performing sorting. Therefore, there is no sorted data (using sort 704) available as a starting point for reordering.

[0148] In conjunction with comparison component 214 and / or sorting component 216, weighting component 218 can generate weights 714 for data layer 340 corresponding to known sample 302. These weights 714 are generated based on the aggregated similarity between encoded known molecular data 328 and encoded target molecular data 628, and between target neutral missing data 608 and known neutral missing data 308. Such weights 714 may be based solely on the comparison of aggregated data, and / or sorting 704 and / or re-sorting 708 may be incorporated into the adopted calculations and / or algorithms.

[0149] In other words, weighting can be based on sorting and reordering, based solely on reordering, and / or alternatives to sorting / reordering.

[0150] The example weight 714 can range from 0 to 1, where 1 indicates high accuracy and 0 indicates almost no accuracy, but other suitable ranges can also be used.

[0151] Comparisons, sorting, reordering, and / or weighting can be performed in any suitable order, and / or at least partially simultaneously.

[0152] Accordingly, based on one or more outputs of comparison component 214, sorting component 216, and / or weighting component 218, matching component 220 may generate one or more predicted matches 270. This generation may be based solely on the output of comparison component 214, and / or may employ one or more sorting 704, reordering 708, and / or weighting 714.

[0153] For example, in one or more embodiments, one or more predicted matches 270 may be obtained for different and / or identical identifications. For example, as described above, matching component 220 may output a set of matches 630. That is, in one or more embodiments, a group consisting of two or more predicted matches 270 corresponding to the same identification (such as an identification with lost neutrality) may include one or more of the reordering 704. In one or more embodiments, a group consisting of two or more predicted matches 270 corresponding to the same identification (such as an identification with lost neutrality) may include one or more of the weights 714. In one or more other embodiments, the two or more identifications may be mutually exclusive (e.g., contradictory).

[0154] The following discussion still refers to Figures 7 and 2, and involves a third set of procedures for further executing the trained model 222 to output one or more additional outputs accompanied by one or more prediction matches 270.

[0155] Accompanying one or more predicted matches 270 may be one or more notifications 290 output by notification component 224. Typically, notification component 224 may generate report data including causal data that identifies associations with one or more levels of known sample data 236. For example, in one or more cases, notification component 224 may generate report data including causal data that associates structural features of target sample 602 (e.g., structural level 604) with specific neutral loss data 318 in known neutral loss data 308 that corresponds to at least one or more bits 324 of encoded known molecular data 328. This helps in understanding the target molecular structural data, neutral loss data and / or spectral data and their origins, and / or the reasoning behind any one or more identifications provided by the model.

[0156] For example, in one or more cases, notification 290 may include sorted, weighted, and / or unsorted / unweighted data, accompanied by or separately from one or more reasoning based on correspondences (e.g., correspondences between molecular structure data, neutral loss data, and / or spectral data). This also helps in understanding the target molecular structure data, neutral loss data, and / or spectral data and their origins, and / or the reasoning behind any one or more identifications provided by the model.

[0157] In one or more embodiments, the notification component 224 may generate visualizations of spectra, molecules, and / or fingerprints (e.g., display data that can be displayed on a graphical user interface communicatively coupled to the non-limiting system 200), wherein one or more layers of the spectra, molecules, and / or fingerprints are labeled and / or marked with identification and / or interpretability information (e.g., identification basis).

[0158] In one or more embodiments, model 222, notification component 224, and / or training component 226 may facilitate the generation and / or modification of data layer 340 stored in library data repository 250 and / or other suitable locations (such as for future identification, training, etc. by model 222 and / or target data annotation system 202). In one or more embodiments, such generation and / or modification may include generating label data that associates known neutral missing data 308 with encoded target molecular data 628, for example, in the data layer of known sample 302 and / or target sample 602.

[0159] In one or more embodiments, training component 226 may facilitate feedback evaluation of prediction matches 270 relative to one or more outputs. For example, this may include a user entity (e.g., using a computing device communicatively coupled to non-restrictive system 200) inputting data to request or modify one or more weights of one or more model hyperparameters of one or more trained models 222.

[0160] As a summary of the components and / or their functions described above, reference is now made to Figures 9 and 10, which illustrate a flowchart of an exemplary non-limiting method 900 that facilitates chemical data comparison and target data annotation processes according to one or more exemplary embodiments described herein (such as the non-limiting system 200 of Figure 2). While the non-limiting method 900 is described relative to the non-limiting system 200 of Figure 2, the non-limiting method 900 is also applicable to other systems described herein, such as the non-limiting system 100 of Figure 1. For the sake of brevity, repeated descriptions of similar elements and / or processes employed in the various embodiments are omitted.

[0161] At 902, the non-limiting method 900 may include: encoding the target molecular data of the target sample into a vectorized format by a system (e.g., encoding component 210) to obtain encoded target molecular data.

[0162] At 904, the non-limiting method 900 may include: encoding the data fingerprint corresponding to the target molecule data into a bit vector comprising multiple bits representing the structural level of a known sample by a system (e.g., encoding component 210), thereby obtaining the encoded target molecule data.

[0163] At 906, the non-limiting method 900 may include: determining, by a system (e.g., encoding component 210 and / or processor 206), whether the vectorization format of the encoded target molecule data has been verified, such as by comparing it with a vectorization format used by known chemical data for comparison with the target chemical data. If yes, the non-limiting method 900 may proceed to step 908. If no, the non-limiting method 900 may return to steps 902 and / or 904.

[0164] At 908, the non-limiting method 900 may include: generating target neutral loss data in a non-coded format based on target spectral data corresponding to target molecule data and target sample, by means of a system (e.g., generation component 212).

[0165] At 910, the non-limiting method 900 may include: comparing encoded known molecular data and encoded target molecular data corresponding to known spectral data through a system (e.g., comparison component 214 and / or model 222).

[0166] At 912, the non-limiting method 900 may include: generating a ranking for the one or more known samples, including the known sample, based on a first similarity level between the encoded known spectral data corresponding to a set of one or more known samples and the encoded target molecule data, by means of a system (e.g., ranking component 216 and / or model 222).

[0167] At 914, the non-limiting method 900 may include: comparing known neutral missing data with target neutral missing data of the target sample by means of a system (e.g., comparison component 214 and / or model 222) to obtain a set of one or more possible matches, including predicted matches, between one or more known samples and the target sample.

[0168] At 916, the non-limiting method 900 may include: performing a comparison of known neutral loss data with target neutral loss data via a system (e.g., comparison component 214 and / or model 222), wherein the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

[0169] At 918, the non-limiting method 900 may include: generating a reordering for the group of one or more known samples, including the known sample, based on a second similarity level between the known neutral missing data corresponding to the group of one or more known samples, including the known neutral missing data, and the target neutral missing data corresponding to the target sample, through a system (e.g., sorting component 216 and / or model 222).

[0170] In one or more embodiments, reordering can be performed without performing sorting.

[0171] In one or more embodiments, comparison, sorting, and reordering may be performed in any suitable order, and / or at least partially simultaneously.

[0172] At 920, the non-limiting method 900 may include: applying a second similarity level to the neutral missing data in the known neutral missing data that corresponds to a known bit of the encoded known molecular data, which matches a target bit of the encoded target molecular data, through a system (e.g., sorting component 216 and / or model 222).

[0173] At 922, the non-limiting method 900 may include: generating data-level weights corresponding to known samples by means of a system (e.g., weighting component 218 and / or model 222), wherein the weights are generated based on the aggregated similarity between encoded known molecular data and encoded target molecular data, and between target neutral loss data corresponding to the target sample and known neutral loss data.

[0174] In one or more embodiments, the weighting may be generated based on sorting and reordering, based solely on reordering, and / or may be an alternative to sorting.

[0175] In one or more embodiments, comparisons, sorting, reordering, and weighting can be performed in any suitable order, and / or at least partially simultaneously.

[0176] At 924, the non-limiting method 900 may include: generating a predicted match between encoded target molecule data, also in a vectorized format, and known neutral loss data of a known sample, through a system (e.g., matching component 220 and / or model 222), the known neutral loss data defining the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0177] At 926, the non-limiting method 900 may include: matching known neutral missing data with bits from a plurality of bits by means of a system (e.g., matching component 220 and / or model 222).

[0178] At 928, the non-limiting method 900 may include: generating report data (e.g., notification 290) by a system (e.g., notification component 224) that includes cause data that associates structural features of the target sample with specific neutral loss data corresponding to at least one or more bits of encoded known molecular data in known neutral loss data.

[0179] As a further summary of the components and / or their functions described above, reference is now made to Figures 11 and 12, which illustrate flowcharts of an exemplary non-limiting method 1100 that facilitates the process of chemical data comparison and target data annotation according to one or more exemplary embodiments described herein (such as the non-limiting system 200 of Figure 2). While the non-limiting method 1100 is described relative to the non-limiting system 200 of Figure 2, the non-limiting method 1100 may also be applied to other systems described herein, such as the non-limiting system 100 of Figure 1. For the sake of brevity, repeated descriptions of similar elements and / or processes employed in the various embodiments are omitted.

[0180] At 1102, the non-limiting method 1100 may include: encoding known molecular data of a known sample into a vectorized format by a system (e.g., encoding component 210) to obtain encoded known molecular data.

[0181] At 1104, the non-limiting method 1100 may include: encoding a data fingerprint corresponding to the known molecular data into a bit vector comprising multiple bits representing the structural level of the known sample by a system (e.g., encoding component 210), thereby obtaining the encoded known molecular data.

[0182] At 1106, the non-limiting method 1100 may include: determining, by a system (e.g., encoding component 210 and / or processor 206), whether the vectorization format of the encoded target molecule data has been validated, such as by comparing it with a vectorization format used by known chemical data for comparison with the target chemical data. If yes, the non-limiting method 1100 may proceed to step 1108. If no, the non-limiting method 1100 may return to steps 1102 and / or 1104.

[0183] At 1108, the non-limiting method 1100 may include: generating known neutral loss data in a non-coded format based on known spectral data corresponding to known samples, by means of a system (e.g., generation component 212).

[0184] At 1110, the non-limiting method 1100 may include: generating tag data by a system (e.g., generation component 212) that associates known neutral loss data with encoded known molecular data.

[0185] At 1112, the non-limiting method 1100 may include: generating a data layer by a system (e.g., generation component 212) that includes known molecular data and known neutral loss data, which are at least partially in a vectorized format.

[0186] At 1114, the non-limiting method 1100 may include: storing the data layer at a data storage location used by the machine learning model that performs the generation of the prediction matching by means of a system (e.g., generation component 212 and / or training component 226).

[0187] At 1116, the non-limiting method 1100 may include: training a machine learning model that performs the generation of predictive matching using a set of data layers through a system (e.g., training component 226), the set of data layers including: encoded known molecular data including the encoded known molecular data and corresponding neutral loss data including the known neutral loss data of a set of known samples including the known sample.

[0188] At 1118, the non-limiting method 1100 may include: encoding the target molecular data of the target sample into a vectorized format by a system (e.g., encoding component 210) to obtain encoded target molecular data.

[0189] At 1120, the non-limiting method 1100 may include: generating a predicted match between encoded target molecule data, also in a vectorized format, and known neutral loss data of a known sample, through a system (e.g., matching component 220 and / or model 222), the known neutral loss data defining the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0190] Additional Invention Content For the sake of brevity, the computer-implemented and non-computer-implemented methods provided herein are depicted and / or described as a series of actions. It should be understood that the invention is not limited to the actions and / or the order of actions shown; for example, actions may occur in one or more orders and / or simultaneously, and may occur alongside other actions not presented or described herein. Furthermore, not all actions shown can be used to implement the computer-implemented and non-computer-implemented methods according to the subject matter. Additionally, the computer-implemented and non-computer-implemented methods may alternatively be represented by a series of interrelated states via state diagrams or events. Furthermore, the computer-implemented methods described below and throughout this specification can be stored on an article of manufacture for transfer and assignment to a computer. As used herein, the term "article of manufacture" is intended to encompass any computer program accessible from any computer-readable device or storage medium.

[0191] The system and / or device has been (and / will further) described herein in relation to the interaction between one or more components. Such systems and / or components may include those components or sub-components specified herein, one or more of the specified components and / or sub-components, and / or additional components. Sub-components may be implemented as components communicatively coupled to other components, rather than being contained within a parent component. One or more components and / or sub-components may be combined into a single component that provides aggregate functionality. These components may interact with one or more other components, which, for simplicity, are not specifically described herein but are known to those skilled in the art.

[0192] In summary, the embodiments described herein relate to target sample data annotation. A system may include a memory and a processor, the memory storing computer-executable components and the processor executing the computer-executable components. The computer-executable components may include encoding components 110, 210 and matching components 120, 220, the encoding components encoding target molecular data 126, 610 of target samples 124, 602 into vectorized formats 128, 626 to obtain encoded target molecular data 130, 628; the matching components generating predicted matches 170, 270 between the encoded target molecular data 126, 610 and known neutral missing data 136, 308 of known samples 122, 302, the known neutral missing data 136, 308 defining mass-to-charge ratio differences 138, 316 between the spectral values ​​134, 312 of the known spectral data 132, 306 corresponding to the known samples 122, 302.

[0193] One or more exemplary embodiments disclosed herein can be plug-and-play applied to a single measurement device, multiple measurement devices, the same measurement device using multiple replaceable components (e.g., chromatographic columns), etc., to compare output data with unknown data, known data, and / or standard data. The framework described herein can be executed in an efficient and at least partially automated manner, thereby reducing manual processes, improving accuracy, and providing automated reasoning for predictions. In one or more cases, the identification data obtained using one or more exemplary embodiments can be used to construct a database of known molecules, neutral loss, and / or spectral data.

[0194] Therefore, one or more exemplary embodiments described herein can be implemented within, in conjunction with, and / or coupled to scientific measurement devices such as mass spectrometers.

[0195] In fact, given one or more exemplary embodiments described herein, the practical application of one or more systems, computer implementation methods, and / or computer program products described herein lies in their ability to generate inference and / or neutral loss data corresponding to a target sample, which cannot be obtained simply by comparing the molecular data and / or neutral loss data of the target sample with a library of known molecular data and / or known neutral loss data. That is, spectral data based on defined spectra may exhibit one or more neutral losses, while other neutral losses may not be apparent. In other words, such unmanifested neutral losses may not occur during fragmentation due to chemical structure (e.g., bond type), chemical properties, failure to reach fragmentation energy, etc. In other words, unmanifested neutral losses may correspond to ions that did not fragment and separate from the target sample for the same and / or different reasons (e.g., chemical structure, chemical properties, fragmentation energy).

[0196] In other words, compared to existing frameworks that cannot provide this functionality, one or more exemplary embodiments described herein can be employed to utilize relevant information about one or more compounds that are different from the target compound to be annotated. For example, predictions can be made regarding neutral loss identification (whether it appears or does not appear in the spectral data), ion identification based on chemical structure (e.g., chemical bond type), and / or target compound identification (but not limited to these predictions) based on molecular structure data, neutral loss data, and / or spectral data (but not limited to these) corresponding to known compounds different from the target compound. For example, such known compounds may belong to the same family, chemical class, etc., as the target compound, and / or may have one or more structural features, ions, and / or neutral losses identical to the target compound. This prediction can be achieved by utilizing a database containing hundreds, thousands, tens of thousands, or more sets of chromatographic data, labeled peaks, etc. (but not limited to these).

[0197] In view of the above advantages, these beneficial effects and / or features are practical applications of computers, thereby providing enhanced (e.g., improved and / or optimized) spectroscopic data analysis. In summary, such computerized tools can constitute concrete and tangible technological improvements in the field of materials analysis, and particularly in the analysis of scientific measurement equipment outputs (such as, but not limited to, the field of spectroscopy).

[0198] Furthermore, based on the disclosed teachings, one or more exemplary embodiments described herein can be employed in real-world systems. For example, one or more embodiments may employ one or more such data layers (e.g., including molecular structure data, neutral loss data, and / or spectral data) to compare the ion recognition, molecular structure, spectral peaks, neutral loss values ​​(e.g., gaps between spectral peaks), etc., of one or more target compounds and / or known compounds. Such comparisons can be used to annotate unknown and / or target compound data and / or generate databases of data layers. Such comparisons can be achieved using databases comprising hundreds, thousands, tens of thousands, or more sets of data layers (but are not limited to this number).

[0199] Furthermore, compared to existing frameworks, this comparison allows for a more comprehensive understanding of the target spectral data. For example, one or more structural and / or neutral loss features can be predicted using a model (such as an artificial intelligence (AI) model or a machine learning (ML) model); this model utilizes the aforementioned database and has learned the correspondences between the molecular structural data, neutral loss data, and / or spectral data included in the database. One or more identified peaks, features, ions, neutral losses, etc., related to known or unknown compounds can be predicted, each corresponding to one or more predicted outputs (such as presented in a sorted and / or weighted format). In one or more cases, the sorted and / or weighted data may be accompanied by one or more sorting and / or weighting rationales based on the correspondences (e.g., correspondences between molecular structural data, neutral loss data, and / or spectral data), and / or these rationales may be provided separately. This helps in understanding the target molecular structural data, neutral loss data, and / or spectral data and their origins, and / or the reasoning behind any one or more identifications provided by the model. In short, the embodiments disclosed herein can therefore improve scientific instrument technology (e.g., improve computer technology that supports such scientific instruments, etc.).

[0200] Furthermore, one or more exemplary embodiments described herein enable a degree of operational scale. For example, spectroscopic data (e.g., target sample data 246) corresponding to two or more compounds (e.g., target sample 602) can be evaluated at least partially in parallel with respect to the same and / or different systems, measuring devices, known chemical data databases, etc.

[0201] The system and / or device has been (and / or will be further) described herein in relation to the interaction between one or more components. Such systems and / or components may include those components or sub-components specified herein, one or more of the specified components and / or sub-components, and / or additional components. Sub-components may be implemented as components communicatively coupled to other components, rather than being contained within a parent component. One or more components and / or sub-components may be combined into a single component that provides aggregation functionality. For example, as described above, in one or more embodiments, model 222 may include one or more of a matching component 220, a weighting component 218, a sorting component 216, and / or a comparison component 214, and / or perform one or more functions described as included therein. These components may interact with one or more other components, which, for simplicity, are not specifically described herein but are known to those skilled in the art.

[0202] One or more exemplary embodiments described herein may be inherently and / or inextricably linked to computer technology in one or more cases and cannot be implemented outside of a computing environment. For example, one or more processes performed by one or more exemplary embodiments described herein may provide program and / or program instruction execution more efficiently and even more practically than existing systems and / or techniques using molecular network generation and / or visualization, such as comparisons with measurement device outputs (e.g., measurement device applications for materials analysis). Systems, computer implementations, and / or computer program products that provide the execution of these processes have great utility in the field of materials analysis and cannot be reasonably and practically implemented outside of a computing environment.

[0203] One or more exemplary embodiments described herein can employ hardware and / or software to solve problems that are highly technical, non-abstract, and cannot be performed by humans as a set of mental actions. For example, one person, or even thousands, cannot efficiently, accurately, and / or effectively analyze computer data / metadata (e.g., defining spectroscopic data) that define fragmented ion mass-to-charge ratios, intensities, inferred neutrality loss, etc., derived from one or more measuring devices, and / or generate digital visual displays quantifying similarities and / or differences between chemical datasets, a process that one or more exemplary embodiments described herein can provide. Furthermore, neither the human brain nor humans with only pen and paper can perform one or more of these processes as implemented according to one or more exemplary embodiments described herein.

[0204] In one or more exemplary embodiments, the processes described herein may be implemented by one or more dedicated computers (e.g., dedicated processing units, dedicated classical computers, dedicated quantum computers, dedicated hybrid classical / quantum systems, and / or another type of dedicated computer) to perform prescribed tasks related to one or more of the aforementioned technologies. The one or more exemplary embodiments described herein and / or components thereof may be used to address new problems arising from the technological advancements mentioned above, the adoption of quantum computing systems, cloud computing systems, computer architectures, and / or other technologies.

[0205] One or more exemplary embodiments described herein may be fully operable to perform one or more other functions (e.g., fully powered on, fully executed, and / or another function), while also performing one or more of the one or more operations described herein.

[0206] To provide additional details about the invention, a list of embodiments and their features is provided below.

[0207] A system includes: a memory storing computer-executable components; and a processor executing the computer-executable components stored in the memory, wherein the computer-executable components include: an encoding component that encodes target molecular data of a target sample into a vectorized format to obtain encoded target molecular data; and a matching component that generates a predictive match between the encoded target molecular data and known neutral loss data of a known sample, the known neutral loss data defining a mass-to-charge ratio difference between spectral values ​​of known spectral data corresponding to the known sample.

[0208] According to the system described in the preceding paragraph, the known neutral loss data includes neutral losses represented by, but not defined by, the spectra corresponding to the known spectral data.

[0209] According to any of the preceding paragraphs, the encoding component encodes the data fingerprint corresponding to the target molecule data into a bit vector comprising a plurality of bits representing the structural level of the target sample, thereby obtaining the encoded target molecule data, and wherein the matching component matches the known neutral missing data with bits among the plurality of bits.

[0210] According to any of the preceding paragraphs, the computer-executable component further includes: a comparison component that compares encoded known molecular data corresponding to the known spectral data with the encoded target molecular data, and compares the known neutral loss data with the target neutral loss data of the target sample to obtain a set of one or more possible matches between one or more known samples and the target sample, including the predicted match.

[0211] According to any of the preceding paragraphs, the computer-executable component further includes: a sorting component that generates a sort for the one or more known samples, including the known samples, based on a first similarity level between the encoded known molecular data corresponding to a set of one or more known samples and the encoded target molecular data.

[0212] According to any of the preceding paragraphs, the sorting component further generates a reordering for the set of one or more known samples, including the known samples, based on a second similarity level between the known neutral missing data corresponding to the set of one or more known samples, including the known neutral missing data, and the target neutral missing data corresponding to the target sample.

[0213] According to any of the preceding paragraphs, the computer-executable component further includes: a generation component that generates the target neutral loss data in a non-encoded format based on target spectral data corresponding to the target molecule data and the target sample.

[0214] According to any of the preceding paragraphs, the second similarity level is applied to the neutral missing data in the known neutral missing data that corresponds to a known bit in the encoded known molecular data, the known bit matching a target bit in the encoded target molecular data.

[0215] According to any of the preceding paragraphs, the computer-executable component further includes: a weighting component that generates data-level weights corresponding to the known sample, wherein the weights are generated based on the aggregated similarity between the encoded known molecular data and the encoded target molecular data, and between the target neutral loss data corresponding to the target sample and the known neutral loss data.

[0216] According to any of the preceding paragraphs, the computer-executable component further includes: a notification component that generates report data including cause data, the cause data associating structural features of the target sample with specific neutral loss data in the known neutral loss data that corresponds to at least one or more bits of the encoded known molecular data.

[0217] A computer-implemented method includes: encoding target molecular data of a target sample into a vectorized format via a system operably coupled to a processor to obtain encoded target molecular data; and generating a predictive match between the encoded target molecular data and known neutral missing data of a known sample via the system, wherein the known neutral missing data defines a mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0218] According to any of the preceding paragraphs, the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

[0219] The computer implementation method according to any of the preceding paragraphs further includes: encoding the known molecular data of the known sample into a vectorized format using the system to obtain encoded known molecular data; and generating the known neutral loss data in a non-encoded format using the system based on the known spectral data corresponding to the known sample.

[0220] The computer implementation method according to any of the preceding paragraphs further includes: encoding a data fingerprint corresponding to the known molecular data into a bit vector by means of the system, the bit vector including multiple bits representing the structural level of the known sample, to obtain the encoded known molecular data.

[0221] The computer implementation method according to any of the preceding paragraphs further includes: generating, through the system, tag data that associates the known neutral loss data with the encoded known molecular data.

[0222] The computer implementation method according to any of the preceding paragraphs further includes: generating a data layer by the system, the data layer including the known molecular data and the known neutral loss data, which are at least partially in the vectorized format; and storing the data layer by the system in a data storage location used by a machine learning model that performs the generation of the predictive matching.

[0223] The computer implementation method according to any of the preceding paragraphs further includes: training the generated machine learning model for performing the prediction matching using a set of data layers through the system, the set of data layers including: encoded known molecular data including the encoded known molecular data and corresponding neutral loss data including the known neutral loss data of a set of known samples including the known samples.

[0224] A computer program product facilitating the process of annotating target samples includes a computer-readable storage medium embodying program instructions, which are executable by a processor to cause the processor to: encode target molecular data of the target samples into a vectorized format to obtain encoded target molecular data; and generate a predictive match between the encoded target molecular data and known neutral missing data of a known sample, wherein the known neutral missing data defines a mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

[0225] According to any of the preceding paragraphs, the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

[0226] According to any of the preceding paragraphs, the computer program product wherein the program instructions are further executable by the processor to cause the processor to: encode a data fingerprint corresponding to the target molecule data into a bit vector comprising a plurality of bits representing the structural level of the target sample, thereby obtaining the encoded target molecule data; and match the known neutral missing data with bits among the plurality of bits.

[0227] Example runtime environment Figure 13 is a schematic block diagram of a runtime environment 1300 to which the described subject can interact. The runtime environment 1300 includes one or more remote components 1310. Remote components 1310 can be hardware and / or software (e.g., threads, processes, computing devices). In one or more exemplary embodiments, the remote component 1310 can be a distributed computer system connected via a communication framework 1340 to local auto-scaling components and / or programs using distributed computer system resources. The communication framework 1340 can include wired network devices, wireless network devices, mobile devices, wearable devices, radio access network devices, gateway devices, femtocellular devices, servers, etc.

[0228] The runtime environment 1300 also includes one or more local components 1320. Local components 1320 may be hardware and / or software (e.g., threads, processes, computing devices). In one or more exemplary embodiments, local components 1320 may include auto-expansion components and / or programs that communicate with / use remote resources such as remote resources 1310 and 1320 connected to the remote distributed computing system via communication framework 1340.

[0229] One possible form of communication between remote component 1310 and local component 1320 could be a data packet format adapted for transmission between two or more computer processes. Another possible form of communication between remote component 1310 and local component 1320 could be a circuit-switched data format adapted for transmission in wireless time slots between two or more computer processes. The operating environment 1300 includes a communication framework 1340 that can be used to facilitate communication between remote component 1310 and local component 1320, and may include an air interface, such as a UMTS network interface via an LTE network. Remote component 1310 can be operatively connected to one or more remote data repositories 1350, such as hard disk drives, solid-state drives, Subscriber Identity Module (SIM) cards, electronic SIMs (eSIMs), device memory, etc., which can be used to store information on the remote component 1310 side of the communication framework 1340. Similarly, local component 1320 can be operatively connected to one or more local data repositories 1330, which can be used to store information on the local component 1320 side of communication framework 1340.

[0230] Example computing environment To provide additional context for the various embodiments described herein, Figure 14 and the following discussion are intended to provide a brief, general description of a suitable computing environment 1400 in which various embodiments of the embodiments described herein may be implemented. Although the embodiments have been described above in the general context of computer-executable instructions that can run on one or more computers, those skilled in the art will recognize that these embodiments may also be implemented in combination with other program modules and / or as a combination of hardware and software.

[0231] Typically, program modules include routines, programs, components, data structures, etc., that perform tasks or implement abstract data types. Furthermore, these methods can be implemented in conjunction with other computer system configurations, including single-processor or multi-processor computer systems, minicomputers, mainframes, Internet of Things (IoT) devices, distributed computing systems, and personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, each of which can be operatively coupled to one or more related devices.

[0232] The embodiments illustrated herein can also be practiced in a distributed computing environment, where certain tasks are performed by remote processing devices linked via a communication network. In a distributed computing environment, program modules can reside on both local and remote memory storage devices.

[0233] Computing devices typically include various media, which may include computer-readable storage media, machine-readable storage media, and / or communication media, these terms being used differently from each other herein. A computer-readable storage media or a machine-readable storage media can be any available storage medium accessible to a computer, including volatile and non-volatile media, removable and non-removable media. By way of example and not limitation, a computer-readable storage media or a machine-readable storage media can be implemented in conjunction with any information storage method or technology, such as computer-readable or machine-readable instructions, program modules, structured data, or unstructured data.

[0234] Computer-readable storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), electrically erasable read-only memory (EEPROM), flash memory or other memory technologies, read-only optical disc memory (CDROM), digital versatile optical disc (DVD), Blu-ray disc (BD) or other optical disc storage, cassette tape, magnetic tape, disk storage or other magnetic storage devices, solid-state drives or other solid-state storage devices, or other tangible and / or non-transient media that can be used to store desired information. In this regard, when the terms “tangible” or “non-transient” are used as modifiers to describe storage media, memories, or computer-readable media herein, only transient signals in transit themselves are excluded, without waiving the rights to all standard storage devices, memories, or computer-readable media that are not solely transient signals in transit themselves.

[0235] Computer-readable storage media can be accessed by one or more local or remote computing devices, for example, through access requests, queries or other data retrieval protocols, to perform various operations on the information stored on the media.

[0236] Communication media typically embody computer-readable instructions, data structures, program modules, or other structured or unstructured data in data signals (such as modulated data signals, like carrier waves or other transmission mechanisms), and include any medium for information transmission or delivery. The term "modulated data signal" or a signal refers to a signal whose one or more characteristics are set or altered in such a way that information is encoded in one or more signals. By way of example and not limitation, communication media include wired media (such as wired networks or direct wired connections) and wireless media (such as acoustic, RF, infrared, and other wireless media).

[0237] Referring again to FIG. 14, an exemplary computing environment 1400 that can implement one or more exemplary embodiments described herein includes a computer 1402, which includes a processing unit 1404, a system memory 1406, and a system bus 1408. The system bus 1408 couples system components (including, but not limited to, the system memory 1406) to the processing unit 1404. The processing unit 1404 can be any of a variety of commercially available processors. A dual-microprocessor or other multiprocessor architecture may also be used as the processing unit 1404.

[0238] System bus 1408 can be any of several types of bus architectures, which can further interconnect with memory buses (with or without a memory controller), peripheral buses, and local buses using any of the various commercially available bus architectures. System memory 1406 includes ROM 1410 and RAM 1412. The basic input / output system (BIOS) can be stored in non-volatile memory, such as ROM, erasable programmable read-only memory (EPROM), or EEPROM, where the BIOS includes basic routines that facilitate the transfer of information between components within computer 1402, such as during startup. RAM 1412 may also include high-speed RAM, such as static RAM for caching data.

[0239] Computer 1402 further includes an internal hard disk drive (HDD) 1414 (e.g., EIDE, SATA) and may include one or more external storage devices 1416 (e.g., floppy disk drive (FDD) 1416, memory stick or flash drive reader, memory card reader, etc.). Although the internal HDD 1414 is shown as being located inside computer 1402, the internal HDD 1414 may also be configured to be externally used in a suitable enclosure (not shown). Additionally, although not shown in computing environment 1400, a solid-state drive (SSD) may be used in addition to or as an alternative to the HDD 1414.

[0240] Other built-in or external storage may include at least one other storage device 1420 with storage medium 1422 (e.g., solid-state storage device, non-volatile storage device, and / or optical disc drive that can be read from or written to removable media such as CD-ROM, DVD, BD, etc.). External storage 1416 may be facilitated by a network virtual machine. HDD 1414, external storage device 1416, and storage device (e.g., drive) 1420 may be connected to system bus 1408 via HDD interface 1424, external storage interface 1426, and drive interface 1428, respectively.

[0241] The drive and its associated computer-readable storage medium provide non-volatile storage of data, data structures, computer-executable instructions, etc. For computer 1402, the drive and storage medium are capable of storing any data in a suitable digital format. Although the description of computer-readable storage media above refers to a specific type of storage device, other types of computer-readable storage media, whether existing or developed in the future, may also be used in the example operating environment, and further, any such storage medium may contain computer-executable instructions for performing the methods described herein.

[0242] Numerous program modules can be stored in the drive and RAM 1412, including the operating system 1430, one or more application programs 1432, other program modules 1434, and program data 1436. All or part of the operating system, applications, modules, and / or data may also be cached in RAM 1412. The systems and methods described herein can be implemented using various commercially available operating systems or combinations of operating systems.

[0243] Computer 1402 may optionally include emulation technology. For example, a virtual machine monitor (not shown) or other intermediate layer may emulate the hardware environment of operating system 1430, and the emulated hardware may optionally differ from the hardware shown in FIG14. In this embodiment, operating system 1430 may include one of a plurality of virtual machines (VMs) hosted on computer 1402. Furthermore, operating system 1430 may provide a runtime environment for application 1432, such as the Java Runtime Environment or the .NET Framework. A runtime environment is a consistent execution environment that allows application 1432 to run on any operating system that includes a runtime environment. Similarly, operating system 1430 may support containers, and application 1432 may be in the form of a container, wherein the container is a lightweight, standalone executable software package that includes, for example, the application's code, runtime, system tools, system libraries, and settings.

[0244] Furthermore, computer 1402 can be configured with security modules, such as a Trusted Processing Module (TPM). For example, with the help of a TPM, the boot component hashes the next boot component over time and waits for the result to match a security value before loading the next boot component. This process can occur at any layer of the code execution stack of computer 1402, for example, at the application execution level or the operating system (OS) kernel level, thereby achieving security during code execution at any level.

[0245] User entities can input commands and information to computer 1402 through one or more wired / wireless input devices, such as keyboard 1438, touchscreen 1440, and pointing devices (such as mouse 1442). Other input devices (not shown) may include microphones, infrared (IR) remote controls, radio frequency (RF) remote controls or other remote controls, joysticks, virtual reality controllers and / or virtual reality headsets, game controllers, styluses, image input devices (e.g., cameras), gesture sensor input devices, visual motion sensor input devices, emotion or face detection devices, biometric input devices (e.g., fingerprint or iris scanners), etc. These and other input devices are typically connected to processing unit 1404 via input device interface 1444, which may be coupled to system bus 1408, but may also be connected to other interfaces such as parallel ports, IEEE 1394 serial ports, game ports, USB ports, IR interfaces, BLUETOOTH® interfaces, etc.

[0246] Monitor 1446 or other types of display devices can also be connected to system bus 1408 via an interface such as video adapter 1448. In addition to monitor 1446, the computer typically includes other peripheral output devices (not shown), such as speakers, printers, etc.

[0247] Computer 1402 can operate in a networked environment using logical connections via wired and / or wireless communications with one or more remote computers, such as remote computer 1450. Remote computer 1450 can be a workstation, server computer, router, personal computer, laptop computer, microprocessor-based entertainment device, peer-to-peer device, or other common network node, and typically includes many or all of the elements described relative to computer 1402, but for simplicity, only memory / storage device 1452 is shown. The described logical connections include wired / wireless connections to a local area network (LAN) 1454 and / or a larger network (e.g., a wide area network (WAN) 1456). Such LAN and WAN networking environments are common in offices and companies and are advantageous for enterprise-wide computer networks (such as intranets) that can all connect to global communication networks (e.g., the Internet).

[0248] When used in a LAN network environment, computer 1402 can connect to local network 1454 via a wired and / or wireless communication network interface or adapter 1458. Adapter 1458 facilitates wired or wireless communication with LAN 1454, which may also include a wireless access point (AP) configured thereon for communicating with adapter 1458 in wireless mode.

[0249] When used in a WAN network environment, computer 1402 may include modem 1460, or may be connected to a communication server on WAN 1456 via other means (such as via the Internet) for establishing communication via WAN 1456. Modem 1460 (which may be built-in or external and may be a wired or wireless device) may be connected to system bus 1408 via input device interface 1444. In a networked environment, program modules described relative to computer 1402 or parts thereof may be stored in remote memory / storage device 1452. The network connection shown is an example, and other means of establishing inter-computer communication links may be used.

[0250] When used in a LAN or WAN network environment, computer 1402 can access cloud storage systems or other network-based storage systems other than or alternative to the external storage device 1416 described above. Generally, the connection between computer 1402 and the cloud storage system can be established via LAN 1454 or WAN 1456, such as adapter 1458 or modem 1460. When computer 1402 is connected to an associated cloud storage system, external storage interface 1426 can manage the storage provided by the cloud storage system with the aid of adapter 1458 and / or modem 1460, just as it would manage other types of external storage. For example, external storage interface 1426 can be configured to provide access to cloud storage sources as if these sources were physically connected to computer 1402.

[0251] Computer 1402 is operable to communicate with any wireless device or entity operatively configured to conduct wireless communication, such as a printer, scanner, desktop and / or laptop computer, portable data assistant, communications satellite, any device or location associated with a wirelessly detectable tag (e.g., a kiosk, newsstand, store shelf, etc.), and telephone. This can include Wi-Fi and BLUETOOTH® wireless technologies. Therefore, communication can be within a defined structure like an existing network, or simply ad hoc communication between at least two devices.

[0252] Other information The embodiments described herein may relate to one or more of systems, methods, apparatuses, and / or computer program products at any possible level of technical detail integration. A computer program product may include a computer-readable storage medium (or media) having computer-readable program instructions thereon to cause a processor to perform aspects of one or more exemplary embodiments described herein. A computer-readable storage medium may be a tangible device capable of retaining and storing instructions for use by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, superconducting storage devices, and / or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of computer-readable storage media may also include: portable computer floppy disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable optical disc read-only storage media (CD-ROM), digital versatile optical disc (DVD), memory sticks, floppy disks, mechanical encoding devices such as punched cards or recessed protrusions on which instructions are recorded, and / or any suitable combination of the foregoing. As used herein, a computer-readable storage medium must not be interpreted as a transient signal itself, such as radio waves and / or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides and / or other transmission media (e.g., light pulses passing through optical cables), and / or electrical signals transmitted through metallic wires.

[0253] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to a corresponding computing / processing device and / or downloaded via a network (e.g., the Internet, a local area network, a wide area network, and / or a wireless network) to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to a computer-readable storage medium within the corresponding computing / processing device. The computer-readable program instructions used to perform the operations of one or more exemplary embodiments described herein may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, integrated circuit configuration data, and / or source code and / or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​(such as Smalltalk, C++, etc.) and / or procedural programming languages ​​(such as the "C" programming language and / or similar programming languages). Computer-readable program instructions can be executed as a standalone software package, entirely on a computer, partially on a computer, partially on a computer and / or partially on a remote computer, or entirely on a remote computer and / or a server. In the latter case, the remote computer can be connected to the computer via any type of network, including local area networks (LANs) and / or wide area networks (WANs), and / or can be connected to an external computer (e.g., via the Internet through an Internet service provider). In one or more exemplary embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGAs), and / or programmable logic arrays (PLAs) can be personalized by utilizing state information from the computer-readable program instructions to execute the computer-readable program instructions intended to perform aspects of one or more exemplary embodiments described herein.

[0254] Aspects of one or more exemplary embodiments described herein are described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to one or more exemplary embodiments described herein. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions. These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, and / or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, can create a manner for implementing the functions / actions specified in the flowchart illustrations and / or block diagram blocks or blocks. These computer-readable program instructions can also be stored in a computer-readable storage medium that can instruct a computer, programmable data processing apparatus, and / or other apparatus to operate in a particular manner, such that a computer-readable storage medium in which the instructions are stored can constitute an article of manufacture containing instructions capable of implementing aspects of the functions / actions specified in the flowchart illustrations and / or block diagram blocks and / or blocks. Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus and / or other apparatus to cause a series of operations to be performed on the computer, other programmable apparatus and / or other apparatus to produce a computer-implemented process, such that the instructions that execute on the computer, other programmable apparatus and / or other apparatus implement the functions / actions specified in the flowchart and / or block diagram blocks and / or blocks.

[0255] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and / or operation of possible implementations of systems, computer-implementable methods, and / or computer program products according to one or more exemplary embodiments described herein. In this respect, each block in a flowchart or block diagram may represent a module, segment, and / or portion of instructions comprising one or more executable instructions for implementing a specified logical function. In one or more alternative implementations, the functions indicated in the blocks may not occur in the order shown in the drawings. For example, two consecutively shown blocks may be executed substantially simultaneously, and / or these blocks may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block and / or combination of blocks in the block diagrams and / or flowcharts may be implemented by a system based on dedicated hardware capable of performing the specified functions and / or actions, and / or implementing one or more combinations of dedicated hardware and / or computer instructions.

[0256] Although the subject matter has been described above within the general context of computer-executable instructions for computer program products running on computers and / or multiple computers, those skilled in the art will recognize that one or more exemplary embodiments herein can also be implemented, at least in part, in parallel with one or more other program modules. Typically, program modules include routines, programs, components, and / or data structures that perform specific tasks and / or implement specific abstract data types. Furthermore, the computer implementation methods described above can be implemented in conjunction with other computer system configurations, including single-processor and / or multi-processor computer systems, small computing devices, mainframe computers, handheld computing devices (e.g., PDAs, telephones), and / or microprocessor-based or programmable consumer and / or industrial electronic products. The aspects shown can also be implemented in a distributed computing environment, where tasks are performed by remote processing devices linked via a communication network. However, one or more aspects (if not all) of one or more exemplary embodiments described herein can be implemented on a standalone computer. In a distributed computing environment, program modules can reside in both local memory storage devices and remote memory storage devices.

[0257] As used herein, the terms “component,” “system,” “platform,” and / or “interface” may refer to and / or include computer-related entities or entities associated with an operational machine having one or more specific functions. Entities described herein may be hardware, a combination of hardware and software, software, or software in execution. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, and / or a computer. As an example, an application running on a server and the server itself can both be components. One or more components may reside in a process and / or an execution thread, and components may reside on a single computer and / or be distributed across two or more computers. In another example, a corresponding component may be executed from various computer-readable media on which various data structures are stored. These components may communicate via local and / or remote processes, such as by signals having one or more data packets (e.g., data from a component that interacts with another component in a local system, a distributed system, and / or interacts with other systems via signals across a network such as the Internet). As another example, a component may be a device having specific functions provided by mechanical parts operated by electrical or electronic circuitry, operated by software and / or firmware applications executed by a processor. In this context, the processor can be located internally and / or externally to the device and can execute at least a portion of the software and / or firmware applications. As yet another example, a component can be a means of providing specific functionality through electronic components without mechanical parts, wherein the electronic components can include a processor and / or other means for executing software and / or firmware that at least partially endow the electronic components with functionality. In one aspect, the component can simulate an electronic component via a virtual machine (e.g., within a cloud computing system).

[0258] Furthermore, the term "or" is intended to mean inclusive "or" rather than exclusive "or". That is, unless otherwise specified or clearly apparent from the context, "X adopts A or B" means any naturally inclusive permutation. That is, if X adopts A; X adopts B; or X adopts both A and B, then "X adopts A or B" is satisfied in any of the foregoing cases. Furthermore, unless otherwise specified or clearly apparent from the context that a singular form is involved, the article "a" as used in the subject matter specification and accompanying drawings should generally be interpreted as meaning "one or more". As used herein, the terms "example" and / or "exemplary" are used to indicate that something is used as an example, instance, or illustration. For the avoidance of doubt, the subject matter described herein is not limited to such examples. Moreover, any aspect or design described herein as an "example" and / or "exemplary" is not necessarily construed as preferred or superior to other aspects or designs, nor does it imply exclusion of equivalent exemplary structures and techniques known to those skilled in the art.

[0259] As used herein, the term "processor" can refer to virtually any computing processing unit and / or device, including but not limited to a single-core processor; a single processor with software multithreading capabilities; a multi-core processor; a multi-core processor with software multithreading capabilities; a multi-core processor with hardware multithreading technology; a parallel platform; and / or a parallel platform with distributed shared memory. Furthermore, a processor can refer to an integrated circuit, application-specific integrated circuit (ASIC), digital signal processor (DSP), field-programmable gate array (FPGA), programmable logic controller (PLC), complex programmable logic device (CPLD), discrete gate or transistor logic, discrete hardware components, and / or any combination thereof, designed to perform the functions described herein. Additionally, processors can utilize nanoscale architectures, such as, but not limited to, molecular and quantum dot-based transistors, switches, and / or gates, to optimize space utilization and / or enhance the performance of the associated device. Processors can be implemented as a combination of computing processing units.

[0260] In this document, terms such as “repository,” “memory,” “data repository,” “data storage,” “database,” and any other information storage component related to the operation and function of the component are used to refer to a “memory component,” an entity embodied in “memory,” or a component containing memory. The memory and / or memory component described herein can be volatile or non-volatile memory, or may include both. By way of illustration and not limitation, non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM), flash memory, and / or non-volatile random access memory (RAM) (e.g., ferroelectric RAM (FeRAM)). Volatile memory can include RAM, for example, which can act as an external cache memory. By way of illustration and not limitation, RAM can be available in a variety of forms, such as Synchronous RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), Direct Rambus RAM (DRRAM), Direct Rambus Dynamic RAM (DRDRAM), and / or Rambus Dynamic RAM (RDRAM). Furthermore, the memory components of the systems and / or computer implementation methods described herein are intended to include, but are not limited to, these and / or any other suitable types of memory.

[0261] The foregoing descriptions are merely examples of systems and computer implementation methods. It is certainly impossible to describe every conceivable combination of components and / or computer implementation methods in order to describe one or more exemplary embodiments, but those skilled in the art will recognize that many further combinations and / or arrangements of one or more exemplary embodiments are possible. Furthermore, the terms “comprising,” “having,” and “possessing” are used to such an extent in the detailed description, claims, appendices, and / or drawings that such terms are intended to be inclusive in a manner similar to the term “comprising,” as interpreted when “comprising” is used as a transitional word in the claims.

[0262] Descriptions of various embodiments may use the phrases “one embodiment,” “multiple embodiments,” “one or more exemplary embodiments,” and / or “some embodiments,” each of which may refer to one or more of the same or different embodiments.

[0263] Various embodiments have been described for illustrative purposes, but these descriptions are not intended to be exhaustive or limited to the embodiments described herein. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, and / or technical improvements to existing technologies in the market, and / or to enable others skilled in the art to understand the embodiments described herein.

Claims

1. A system comprising: Memory, which stores computer-executable components; and A processor that executes the computer-executable component stored in the memory, wherein the computer-executable component includes: The encoding component encodes the target molecular data of the target sample into a vectorized format, producing encoded target molecular data. and A matching component generates a predicted match between the encoded target molecule data and known neutral missing data of a known sample, wherein the known neutral missing data is defined as the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

2. The system of claim 1, wherein the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

3. The system of claim 1, wherein the encoding component encodes the data fingerprint corresponding to the target molecule data into a bit vector comprising multiple bits representing the structural level of the target sample, thereby obtaining the encoded target molecule data, and The matching component matches the known neutral missing data with bits from the plurality of bits.

4. The system of claim 1, wherein the computer-executable component further comprises: The comparison component compares the encoded known molecular data corresponding to the known spectral data with the encoded target molecular data, and compares the known neutral loss data with the target neutral loss data of the target sample to obtain one or more possible matches between one or more known samples and the target sample, including the predicted matches.

5. The system of claim 1, wherein the computer-executable component further comprises: A ranking component generates a ranking for the one or more known samples, including the known samples, based on a first similarity level between the encoded known molecular data corresponding to a set of one or more known samples and the encoded target molecular data.

6. The system of claim 5, wherein the sorting component further generates a reordering for the set of one or more known samples, including the known samples, based on a second similarity level between the known neutral missing data corresponding to the set of one or more known samples, including the known neutral missing data, and the target neutral missing data corresponding to the target sample.

7. The system of claim 6, wherein the computer-executable component further comprises: A generation component generates the target neutral loss data in a non-encoded format based on target spectral data corresponding to the target molecule data and the target sample.

8. The system of claim 6, wherein the second similarity level is applied to the neutral missing data in the known neutral missing data that corresponds to a known bit of the encoded known molecular data, the known bit matching a target bit of the encoded target molecular data.

9. The system of claim 1, wherein the computer-executable component further comprises: A weighting component that generates data-level weights corresponding to the known samples, wherein the weights are generated based on the aggregated similarity between the encoded known molecular data and the encoded target molecular data, and between the target neutral missing data corresponding to the target sample and the known neutral missing data.

10. The system of claim 1, wherein the computer-executable component further comprises: A notification component that generates report data including cause data, which associates the structural features of the target sample with specific neutral missing data in the known neutral missing data that corresponds to at least one or more bits of the encoded known molecular data.

11. A computer-implemented method, comprising: The target molecular data of the target sample is encoded into a vectorized format by a system that is operatively coupled to the processor. as well as The system generates a predictive match between the encoded target molecule data and the known neutral missing data of a known sample, wherein the known neutral missing data is defined as the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

12. The computer implementation method of claim 11, wherein the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

13. The computer implementation method according to claim 11, further comprising: The system encodes the known molecular data of the known sample into a vectorized format to obtain the encoded known molecular data. as well as The system generates the known neutral loss data in a non-coded format based on the known spectral data corresponding to the known samples.

14. The computer implementation method according to claim 13, further comprising: The system encodes the data fingerprint corresponding to the known molecular data into a bit vector, the bit vector including multiple bits representing the structural level of the known sample, thus obtaining the encoded known molecular data.

15. The computer implementation method according to claim 13, further comprising: The system generates tag data that associates the known neutral missing data with the encoded known molecular data.

16. The computer implementation method according to claim 13, further comprising: The system generates a data layer that includes the known molecular data and the known neutral loss data, which are at least partially in the vectorized format. as well as The system stores the data layer in a data storage location used by the generated machine learning model that performs matching on the predictions.

17. The computer implementation method according to claim 11, further comprising: The system trains the generated machine learning model that performs the prediction matching using a set of data layers, the set of data layers including: encoded known molecular data including the encoded known molecular data and corresponding neutral loss data including the known neutral loss data.

18. A computer program product for facilitating a process of annotating target samples, the computer program product comprising a computer-readable storage medium embodying program instructions, and the program instructions being executable by a processor to cause the processor to: The processor encodes the target molecule data of the target sample into a vectorized format to obtain the encoded target molecule data; and The processor generates a predicted match between the encoded target molecule data and the known neutral missing data of a known sample, wherein the known neutral missing data is defined as the mass-to-charge ratio difference between the spectral values ​​of the known spectral data corresponding to the known sample.

19. The computer program product of claim 18, wherein the known neutral loss data includes neutral loss represented by, but not defined by, a spectrum corresponding to the known spectral data.

20. The computer program product of claim 18, wherein the program instructions are further executable by the processor to cause the processor to: The data fingerprint corresponding to the target molecule data is encoded into a bit vector comprising multiple bits representing the structural level of the target sample, thereby obtaining the encoded target molecule data; and The processor matches the known neutral missing data with the bits among the plurality of bits.

Images