Systems and methods for diagnosing biological conditions associated with periodic changes in metal metabolism

Abstract

A method for assessing a subject for a biological condition related to metal metabolism, the method comprising sampling locations along a biological sample of the subject to obtain a number of ion samples. Each ion sample corresponds to a location on the biological sample, and each location represents an amount of growth of the biological sample. The obtained ions are analyzed with a mass spectrometer, thereby obtaining a plurality of traces. Each such trace represents a corresponding element isotope of a plurality of element isotopes over time. A set of features is derived from the traces. Each feature is determined from a change in a single isotope or combination of isotopes in the plurality of traces. The set of features is input into a trained classifier to obtain a probability that the subject has the biological condition related to metal metabolism.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to U.S. Provisional Patent Application No. 62 / 858,260, filed June 6, 2019, entitled "Systems and Methods for Hair Based Diagnostics for Autism Spectrum Disorders," which is hereby incorporated by reference. Technical Field

[0003] This disclosure generally relates to the diagnosis of such metal metabolism-related biological conditions by analyzing biological samples from subjects that are being tested for biological conditions. Background Technology

[0004] Metal ions play important roles in many biological processes that are structurally and functionally significant to humans. Imbalances in certain metal ions, due to variations in their quantity in nutrition or metabolic disorders, are associated with numerous biological conditions. Imbalances include both excesses and deficiencies of certain metal ions. Examples of biological conditions associated with metal metabolism include neurological conditions (e.g., autism spectrum disorder, schizophrenia, or attention deficit hyperactivity disorder (ADHD)), neurodegenerative conditions (e.g., amyotrophic lateral sclerosis (ALS), Alzheimer's disease, Parkinson's disease, and Huntington's disease), and some cancers (e.g., childhood cancer).

[0005] Recent research has shown a link between autism spectrum disorder and metabolic dysfunction, specifically metal dysregulation (see, for example, Cheng et al., “Metabolic Dysfunction Underlying Autism Spectrum Disorder and Potential Treatment Approaches,” Front. Molecular Neuroscience, 10, p. 34, February 2017 and Arora et al., “Fetal and postnatal metaldysregulation in Autism,” Nature Communications, 8, p. 15493, June 2017). As another example, recent studies have shown a link between neuronal degeneration and the circadian rhythms of metals detectable in the hair and / or teeth of subjects (see, for example, Appenzeller et al., “Stable Isotope Ratios in Hair and Teeth Reflect Biologic Rhythms,” PLoS ONE 2(7):e636. https: / / doi.org / 10.1371 / journal.pone.0000636, April 2017).

[0006] In light of the above background, there is a need in the art for improved systems and methods for the accurate diagnosis of biological conditions related to metal metabolism. Specifically, there is a need for biomarkers that can be detected by non-invasive methods for the diagnosis of biological conditions related to metal metabolism. Summary of the Invention

[0007] Therefore, there is a need for accurate methods and systems for the diagnosis of biological conditions related to metal metabolism, and especially for non-invasive diagnosis. This disclosure addresses these needs, for example, by providing biomarkers for diagnosing biological conditions related to metal metabolism. The biological samples include human biological samples containing deposits of certain metals and related to growth. Such biological samples can be hair shafts, teeth, and nails. The non-invasive biomarkers of this disclosure can be used for the diagnosis of young children, even infants under one year old.

[0008] According to some embodiments, a method for assessing a subject's first biological symptom related to metal metabolism includes sampling at each corresponding location of a plurality of positions along a reference line on a biological sample related to metal metabolism of the subject, thereby obtaining a plurality of ion samples. Each of the plurality of ion samples corresponds to a different location among the plurality of positions, and each of the plurality of positions represents a different growth stage of the biological sample related to metal metabolism. The method includes analyzing each of the plurality of ion samples (e.g., using mass spectrometry or other spectroscopic methods) thereby obtaining a first dataset comprising a plurality of traces. Each of the plurality of traces is the concentration of a corresponding elemental isotope among a plurality of elemental isotopes determined jointly by the plurality of ion samples over time. The method includes deriving a second dataset comprising a set of features from the plurality of traces. Each corresponding feature in the set of features is determined by the variation of a single isotope or combination of isotopes in the plurality of traces. The method includes inputting the set of features into a trained classifier, thereby obtaining from the trained classifier the probability that the subject suffers from the first biological symptom related to metal metabolism.

[0009] In some embodiments, the plurality of element isotopes are selected from the element isotopes listed in Table 1. In some embodiments, the plurality of element isotopes includes at least 22 element isotopes listed in Table 1.

[0010] According to some embodiments, each feature in the set of features is associated with a single corresponding trace or two corresponding traces among the plurality of traces. In some embodiments, the set of features is selected from the features listed in Table 2, and optionally, the set of features further includes one or more features listed in Table 3. In some embodiments, the set of features includes at least 23 features listed in Table 2.

[0011] In some embodiments, the first biological symptom associated with metal metabolism is selected from the group consisting of: autism spectrum disorder (ADS), attention deficit / hyperactivity disorder (ADHD), amyotrophic lateral sclerosis (ALS), schizophrenia, irritable bowel disease (IBD), pediatric kidney transplant rejection, and pediatric cancer.

[0012] In some embodiments, assessing the subject's first biological symptom related to metal metabolism further includes distinguishing between the first biological symptom related to metal metabolism and a second biological symptom related to metal metabolism, the second biological symptom being different from the first biological symptom related to metal metabolism. In some embodiments, the first biological symptom is autism spectrum disorder, and the second biological symptom is attention deficit / hyperactivity disorder.

[0013] In some embodiments, the subject is a human being. In some embodiments, the subject is less than 1 year old, less than 2 years old, less than 3 years old, less than 4 years old, or less than 5 years old.

[0014] In some embodiments, the metal metabolism-related biological samples of the subject are selected from the group consisting of hair shafts, teeth, and nails.

[0015] In some embodiments, the method further includes pretreating the hair shaft with a solvent and / or irradiating the hair shaft with a low-power laser to remove any debris on the hair shaft before sampling the hair shaft of the subject. In some embodiments, the metal metabolism-related biological sample of the subject is the hair shaft, and the reference line corresponds to the longitudinal direction of the hair shaft. In some embodiments, the metal metabolism-related biological sample of the subject is the tooth, and the reference line corresponds to the neonatal line on the enamel surface of the tooth.

[0016] In some embodiments, the method further includes pretreating the subject's metal metabolism-related biological sample with a solvent or surfactant prior to the sampling. In some embodiments, the method further includes irradiating the subject's metal metabolism-related biological sample with a low-power laser prior to the sampling to remove any debris from the subject's metal metabolism-related biological sample.

[0017] In some embodiments, the sampling includes irradiating the subject's metal metabolism-related biological sample with the laser, thereby extracting a plurality of particles from the subject's metal metabolism-related biological sample, and ionizing the plurality of particles using an inductively coupled plasma mass spectrometer, thereby obtaining the plurality of ion samples.

[0018] In some embodiments, the plurality of locations are arranged in such a sequence that a first location among the plurality of locations along the subject's metal metabolism-related biological sample corresponds to the location closest to the tip of the subject's metal metabolism-related biological sample. In some embodiments, the plurality of locations comprises at least 100, 150, 200, 250, 300, 350, 400, 450, or 500 locations.

[0019] In some embodiments, each of the plurality of traces includes a plurality of data points. Each data point is an instance of a corresponding location among the plurality of locations.

[0020] In some embodiments, deriving the second dataset involves removing such data points from the plurality of data points that do not meet a first criterion. The first criterion includes the mean absolute difference between adjacent data points being three times the standard deviation of the mean absolute difference between adjacent data points.

[0021] In some embodiments, the concentration of the corresponding element isotope corresponds to the relative abundance of the corresponding element isotope relative to a reference element isotope contained in the plurality of ion samples. In some embodiments, the reference element isotope is sulfur.

[0022] In some embodiments, the set of features is selected from average diagonal length, determinism, recursion time, entropy, capture time, and hierarchicality.

[0023] In some embodiments, the trained classifier calculates:

[0024]

[0025] Where p (subject) is the probability that the subject has the first biological symptom related to metal metabolism, e is the Euler number, and α is the probability that β1x1 + ... + β k x k When x equals zero, it is a calculated parameter associated with the probability that the subject suffers from the aforementioned biological symptom related to metal metabolism. 1,…,k The value corresponding to each feature in the set of features, which includes features 1 to k, and β 1,…,k The weight parameters correspond to the weights associated with each of the set of features, which includes features 1 to k.

[0026] In some embodiments, the method further includes determining that the subject suffers from the first biological symptom related to metal metabolism based on determining that p (the subject) is higher than a predetermined threshold.

[0027] In some embodiments, the biological symptoms associated with metal metabolism are related to periodic dysregulation of the metabolism of multiple metals, which correspond to the multiple element isotopes.

[0028] According to some embodiments, an apparatus for assessing a subject's metal metabolism-related biological condition includes one or more processors and a memory storing one or more programs executed by the processors. The programs include instructions to sample at each of a plurality of locations along a reference line on a metal metabolism-related biological sample of the subject, thereby obtaining a plurality of ion samples. Each of the plurality of ion samples corresponds to a different location among the plurality of locations. Each location represents a different growth phase of the metal metabolism-related biological sample. The programs include instructions to analyze each of the plurality of ion samples using a mass spectrometer, thereby obtaining a first dataset comprising a plurality of traces. Each of the plurality of traces is the concentration of a corresponding elemental isotope among multiple elemental isotopes determined jointly by the plurality of ion samples over time. The programs include instructions to derive a second dataset comprising a set of features from the plurality of traces, each corresponding feature being determined by variations in a single isotope or combination of isotopes in the plurality of traces. The one or more procedures include instructions for inputting the set of features into a trained classifier, thereby obtaining from the trained classifier the probability that the subject suffers from the biological symptom related to metal metabolism.

[0029] According to some embodiments, a non-transitory computer-readable storage medium embeds one or more computer programs for classification. The one or more computer programs contain instructions, when executed by a computer system, to perform a method for assessing a subject's metal metabolism-related biological condition. The method includes sampling at each corresponding location among a plurality of locations along a reference line on the subject's metal metabolism-related biological sample, thereby obtaining a plurality of ion samples. Each of the plurality of ion samples corresponds to a different location among the plurality of locations, and each of the plurality of locations represents a different growth stage of the metal metabolism-related biological sample. The method includes analyzing each of the plurality of ion samples with a mass spectrometer, thereby obtaining a first dataset comprising a plurality of traces. Each of the plurality of traces is the concentration of a corresponding elemental isotope among a plurality of elemental isotopes determined jointly by the plurality of ion samples over time. The method includes deriving a second dataset comprising a set of features from the plurality of traces. Each corresponding feature in the set of features is determined by changes in a single isotope or combination of isotopes in the plurality of traces. The method includes inputting the set of features into a trained classifier, thereby obtaining from the trained classifier the probability that the subject has the first biological symptom related to metal metabolism.

[0030] According to some embodiments, a classification method is executed on a computer system having one or more processors and memory storing one or more programs to be executed by the one or more processors. The classification method is executed for each corresponding training subject among a plurality of training subjects. A first subset of the training subjects has a first diagnostic state corresponding to having a first biological symptom related to metal metabolism, and a second subset of the training subjects has a second diagnostic state corresponding to not having the first biological symptom related to metal metabolism. The classification method includes sampling each of a plurality of corresponding positions on a corresponding reference line on a corresponding biological sample related to metal metabolism of the corresponding training subject, thereby obtaining a plurality of corresponding ion samples. Each of the plurality of corresponding ion samples corresponds to a different position among the plurality of corresponding positions. Each of the plurality of corresponding positions represents a different growth stage of the corresponding biological sample related to metal metabolism. The classification method includes analyzing each of the plurality of corresponding ion samples with a mass spectrometer, thereby obtaining a corresponding first dataset containing a plurality of corresponding traces. Each of the plurality of corresponding traces is the concentration of a corresponding element isotope among multiple element isotopes determined jointly by the plurality of corresponding ion samples over time. The classification method includes deriving a corresponding second dataset containing a corresponding set of features from the corresponding plurality of traces. Each corresponding feature in the corresponding set of features is determined by a variation of a single isotope or combination of isotopes in the corresponding plurality of traces. The classification method includes training an untrained or partially untrained classifier to obtain a trained classifier using: (i) the corresponding set of features for each corresponding second dataset of each of the plurality of training subjects; and (ii) a corresponding diagnostic state selected from the first diagnostic state and the second diagnostic state for each of the plurality of training subjects. The classifier provides an indication of whether the test subject suffers from the first biological symptom related to metal metabolism based on the value of a feature from a set of features obtained from a metal metabolism-related biological sample of the test subject.

[0031] In some embodiments, the trained classifier is a neural network algorithm, a support vector machine algorithm, a decision tree algorithm, an unsupervised clustering model algorithm, a supervised clustering model algorithm, or a regression model.

[0032] In some embodiments, the trained classifier is polynomial or binomial. In some embodiments, the plurality of element isotopes are selected from the element isotopes listed in Table 1.

[0033] In some embodiments, each of the set of features is associated with a single corresponding trace or two corresponding traces of the plurality of traces. In some embodiments, the set of features is selected from the features listed in Table 2, and optionally, the set of features further includes one or more features listed in Table 3.

[0034] In some embodiments, the first biological symptom associated with metal metabolism is selected from the group consisting of: autism spectrum disorder (ADS), attention deficit / hyperactivity disorder (ADHD), amyotrophic lateral sclerosis (ALS), schizophrenia, irritable bowel disease (IBD), pediatric kidney transplant rejection, and pediatric cancer.

[0035] In some embodiments, assessing the subject's first biological symptom related to metal metabolism further includes distinguishing between the first biological symptom related to metal metabolism and a second biological symptom related to metal metabolism, the second biological symptom being different from the first biological symptom related to metal metabolism. In some embodiments, the first biological symptom is autism spectrum disorder, and the second biological symptom is attention deficit / hyperactivity disorder.

[0036] In some embodiments, the subject is a human being. In some embodiments, the subject is less than 1 year old, less than 2 years old, less than 3 years old, less than 4 years old, or less than 5 years old.

[0037] In some embodiments, the metal metabolism-related biological samples of the subject are selected from the group consisting of hair shafts, teeth, and nails.

[0038] In some embodiments, the method further includes pretreating the hair shaft with a solvent and / or irradiating the hair shaft with a low-power laser to remove any debris on the hair shaft before sampling the hair shaft of the subject. In some embodiments, the metal metabolism-related biological sample of the subject is the hair shaft, and the reference line corresponds to the longitudinal direction of the hair shaft. In some embodiments, the metal metabolism-related biological sample of the subject is the tooth, and the reference line corresponds to the neonatal line on the enamel surface of the tooth.

[0039] In some embodiments, the method further includes pretreating the subject's metal metabolism-related biological sample with a solvent or surfactant prior to the sampling. In some embodiments, the method further includes irradiating the subject's metal metabolism-related biological sample with a low-power laser prior to the sampling to remove any debris from the subject's metal metabolism-related biological sample.

[0040] In some embodiments, the sampling includes irradiating the subject's metal metabolism-related biological sample with the laser, thereby extracting a plurality of particles from the subject's metal metabolism-related biological sample, and ionizing the plurality of particles using an inductively coupled plasma mass spectrometer, thereby obtaining the plurality of ion samples.

[0041] In some embodiments, the plurality of locations are arranged in such a sequence that a first location among the plurality of locations along the subject's metal metabolism-related biological sample corresponds to the location closest to the tip of the subject's metal metabolism-related biological sample. In some embodiments, the plurality of locations comprises at least 100, 150, 200, 250, 300, 350, 400, 450, or 500 locations.

[0042] In some embodiments, each of the plurality of traces includes a plurality of data points. Each data point is an instance of a corresponding location among the plurality of locations.

[0043] In some embodiments, deriving the second dataset involves removing such data points from the plurality of data points that do not meet a first criterion. The first criterion includes the mean absolute difference between adjacent data points being three times the standard deviation of the mean absolute difference between adjacent data points.

[0044] In some embodiments, the concentration of the corresponding element isotope corresponds to the relative abundance of the corresponding element isotope relative to a reference element isotope contained in the plurality of ion samples. In some embodiments, the reference element isotope is sulfur.

[0045] In some embodiments, the set of features is selected from average diagonal length, determinism, recursion time, entropy, capture time, and hierarchicality.

[0046] In some embodiments, the trained classifier calculates:

[0047]

[0048] Where p (subject) is the probability that the subject has the first biological symptom related to metal metabolism, e is the Euler number, and α is the probability that β1x1 + ... + β k x k When x equals zero, it is a calculated parameter associated with the probability that the subject suffers from the aforementioned biological symptom related to metal metabolism. 1,…,k The value corresponding to each feature in the set of features, which includes features 1 to k, and β 1,…,kThe weight parameters correspond to the weights associated with each of the set of features, which includes features 1 to k.

[0049] In some embodiments, the method further includes determining that the subject suffers from the first biological symptom related to metal metabolism based on determining that p (the subject) is higher than a predetermined threshold.

[0050] In some embodiments, the biological symptoms associated with metal metabolism are related to periodic dysregulation of the metabolism of multiple metals, which correspond to the multiple element isotopes.

[0051] According to some embodiments, a classification device includes one or more processors and a memory storing one or more programs executable by the one or more processors. The one or more programs include instructions for performing a classification method. The classification method is performed for each corresponding training subject among a plurality of training subjects. A first subset of the training subjects has a first diagnostic state corresponding to having a first biological symptom related to metal metabolism, and a second subset of the training subjects has a second diagnostic state corresponding to not having the first biological symptom related to metal metabolism. The classification method includes sampling each of a plurality of corresponding positions on a corresponding reference line on a corresponding biological sample related to metal metabolism of the corresponding training subject, thereby obtaining a plurality of corresponding ion samples. Each of the plurality of corresponding ion samples corresponds to a different position among the plurality of corresponding positions. Each of the plurality of corresponding positions represents a different growth stage of the corresponding biological sample related to metal metabolism. The classification method includes analyzing each of the plurality of corresponding ion samples using a mass spectrometer, thereby obtaining a corresponding first dataset containing a plurality of corresponding traces. Each of the corresponding plurality of traces is the concentration of a corresponding element isotope among multiple elemental isotopes determined over time based on the corresponding plurality of ion samples. The classification method includes deriving a corresponding second dataset containing a corresponding set of features from the corresponding plurality of traces. Each corresponding feature in the corresponding set of features is determined by the variation of a single isotope or combination of isotopes in the corresponding plurality of traces. The classification method includes training an untrained or partially untrained classifier to obtain a trained classifier using: (i) the corresponding set of features for each corresponding second dataset of each of the plurality of training subjects; and (ii) a corresponding diagnostic status selected from the first diagnostic status and the second diagnostic status for each of the plurality of training subjects. The classifier provides an indication of whether the test subject suffers from the first biological symptom related to metal metabolism based on the value of a feature from a set of features obtained from a biological sample related to metal metabolism of the test subject.

[0052] According to some embodiments, a non-transitory computer-readable storage medium embeds one or more computer programs for classification. The one or more computer programs contain instructions, when executed by a computer system, to cause the computer system to perform a classification method. The classification method is performed for each corresponding training subject among a plurality of training subjects. A first subset of the training subjects has a first diagnostic state corresponding to having a first biological symptom related to metal metabolism, and a second subset of the training subjects has a second diagnostic state corresponding to not having the first biological symptom related to metal metabolism. The classification method includes sampling each of a plurality of corresponding positions on a corresponding reference line on a corresponding biological sample related to metal metabolism of the corresponding training subject, thereby obtaining a plurality of corresponding ion samples. Each of the plurality of corresponding ion samples corresponds to a different position among the plurality of corresponding positions. Each of the plurality of corresponding positions represents a different growth stage of the corresponding biological sample related to metal metabolism. The classification method includes analyzing each of the plurality of corresponding ion samples using a mass spectrometer, thereby obtaining a corresponding first dataset containing a plurality of corresponding traces. Each of the corresponding plurality of traces is the concentration of a corresponding element isotope among multiple elemental isotopes determined over time based on the corresponding plurality of ion samples. The classification method includes deriving a corresponding second dataset containing a corresponding set of features from the corresponding plurality of traces. Each corresponding feature in the corresponding set of features is determined by the variation of a single isotope or combination of isotopes in the corresponding plurality of traces. The classification method includes training an untrained or partially untrained classifier to obtain a trained classifier using: (i) the corresponding set of features for each corresponding second dataset of each of the plurality of training subjects; and (ii) a corresponding diagnostic status selected from the first diagnostic status and the second diagnostic status for each of the plurality of training subjects. The classifier provides an indication of whether the test subject suffers from the first biological symptom related to metal metabolism based on the value of a feature from a set of features obtained from a biological sample related to metal metabolism of the test subject.

[0053] As disclosed herein, any embodiments disclosed herein may be applied in any way where applicable.

[0054] Further aspects and advantages of this disclosure will become apparent to those skilled in the art from the following detailed description, in which only illustrative embodiments of the disclosure are shown and described. As will be appreciated, this disclosure is capable of other and different embodiments, and certain details thereof can be modified in various obvious ways without departing from this disclosure. Therefore, the drawings and description are to be regarded in an illustrative rather than restrictive manner. Attached Figure Description

[0055] Figure 1A A block diagram of an example computing device according to some embodiments of the present disclosure is shown.

[0056] Figure 2A A flowchart is provided for a method for assessing biological symptoms of a subject according to some embodiments of the present disclosure.

[0057] Figure 2B Exemplary illustrations of hair samples, tooth samples, and nail samples of subjects according to some embodiments of the present disclosure are provided.

[0058] Figure 2C Exemplary schematic diagrams of laser sampling of hair shafts of a subject according to some embodiments of the present disclosure are provided.

[0059] Figure 2D Exemplary illustrations are provided of traces of the concentration of elemental isotopes over time according to some embodiments of the present disclosure.

[0060] Figure 2E Exemplary illustrations are provided that correspond to features of individual isotope variations derived from traces, according to some embodiments of the present disclosure.

[0061] Figure 2F Illustrations of experimental data for distinguishing autism spectrum disorder from other neurodevelopmental disorders, according to some embodiments of this disclosure, are provided. Figure 2F In this study, cases of autism spectrum disorder (labeled ASD) were compared with cases of attention deficit / hyperactivity disorder (labeled ADHD), subjects diagnosed with a comorbid diagnosis of ASD and ADHD (labeled CM), and neurotypical subjects who had received a diagnosis of no neurodevelopmental disorder (labeled NT).

[0062] Figures 3A-3E Together, flowcharts are provided for processes and characteristics of assessing biological symptoms in subjects according to some embodiments of this disclosure, wherein optional boxes are indicated by dashed boxes.

[0063] Figure 4Flowcharts are provided for processes and characteristics of training a classifier to assess the biological symptoms of a subject, according to some embodiments of the present disclosure, wherein optional boxes are indicated by dashed boxes.

[0064] Figure 5A , 5B Figures 5C and 5D illustrate experimental recipient operating characteristic (ROC) curves for assessing autism spectrum disorder according to some embodiments.

[0065] Figure 6 ROC curves are shown for evaluating the accuracy of the disclosed method for assessing amyotrophic lateral sclerosis (ALS) according to some embodiments.

[0066] Figure 7 ROC curves are shown for evaluating the accuracy of the disclosed method for assessing schizophrenia according to some embodiments.

[0067] Figure 8 ROC curves are shown according to some embodiments for evaluating the accuracy of the disclosed method for assessing irritable bowel syndrome.

[0068] Figure 9 ROC curves are shown for evaluating the accuracy of the disclosed method for assessing kidney transplant rejection according to some embodiments.

[0069] Figure 10 ROC curves are shown according to some embodiments for evaluating the accuracy of the disclosed method for assessing pediatric cancer.

[0070] Throughout the various views in the accompanying drawings, similar reference numerals refer to corresponding parts. The drawings are not drawn to scale. Detailed Implementation

[0071] This disclosure provides systems and methods for assessing metal metabolism-related biological conditions in subjects based on biosamples related to metal metabolism. Specifically, the disclosed methods provide biomarkers that can be obtained noninvasively from the subject. The methods can be used to assess subjects of any age and are particularly suitable for the diagnosis of younger children, even infants under 1 year old, enabling early treatment and intervention.

[0072] definition.

[0073] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in the specification and appended claims of this invention, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to equally encompass the plural forms. It will also be understood that the term “and / or,” as used herein, refers to and includes any and all possible combinations of one or more of the associated listed items. It will be further understood that, when used in this specification, the terms “comprises” and / or “comprising” specify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0074] As used herein, depending on the context, the term "if" can be interpreted as meaning "when," "upon," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if [the stated condition or event] is detected" can be interpreted as meaning "when determination," "in response to determination," "when [the stated condition or event] is detected," or "in response to the detection of [the specified condition or event]."

[0075] As used herein, biological symptoms associated with metal metabolism (also referred to as metal metabolism disorders) are biological symptoms associated with or caused by cyclical dysregulations in the metabolism of certain metals. Cyclic dysregulations can manifest as cyclical decreases (e.g., deficiencies) in the uptake of one or more metals, cyclical increases in the uptake of one or more metals, or a combination of cyclical decreases and increases in the uptake of said one or more metals. Non-limiting examples of biological symptoms associated with metal metabolism include autism spectrum disorder (ADS), attention deficit / hyperactivity disorder (ADHD), amyotrophic lateral sclerosis (ALS), schizophrenia, kidney transplant rejection, certain types of cancer, Alzheimer's disease, Parkinson's disease, Huntington's disease, metabolic disorders (obesity and irritable bowel disease (IBD)) and / or any symptom or condition associated with metal metabolism.

[0076] As used herein, metal metabolism-related biological samples refer to human biological samples (e.g., hair, nails, and teeth) containing deposits of certain metals and associated with growth. The metal metabolism-related biological samples of this disclosure need to express growth along a reference line so that the abundance of certain metal deposits can be detected relative to time. These metal metabolism-related biological samples thus facilitate the detection of periodic changes in the abundance of certain metals. In some embodiments, the metal metabolism-related biological sample comprises a hair shaft, wherein the reference line corresponds to a line along the longitudinal direction of the hair shaft. In some embodiments, the metal metabolism-related biological sample comprises a tooth, wherein the reference line corresponds to a nascent line on the enamel surface of the tooth. In some embodiments, the metal metabolism-related biological sample comprises a nail, wherein the reference line corresponds to a line in the nail growth direction. For example, the reference line extends from the nail root toward the nail tip.

[0077] As used in this paper, the term “trained classifier” refers to a model with specific parameters (weights) and thresholds that is ready to be applied to previously unseen samples (e.g., machine learning algorithms such as logistic regression, neural networks, regression, support vector machines, clustering algorithms, decision trees).

[0078] As used in this paper, the term "untrained classifier or partially trained classifier" refers to a model (e.g., machine learning algorithms such as logistic regression, neural networks, regression, support vector machines, clustering algorithms, decision trees) that has at least some unfixed parameters (weights) and thresholds and is ready to be trained on a training set to optimize and fix the parameters and thresholds.

[0079] It should also be understood that although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, without departing from the scope of this disclosure, a first subject may be referred to as a second subject, and similarly, a second subject may be referred to as a first subject. Although both the first subject and the second subject are subjects, they are not the same subject. Furthermore, the terms "subject," "user," and "patient" are used interchangeably herein.

[0080] As used herein, the term "subject" refers to a person (e.g., a human male, a human female, a fetus, a pregnant woman, a child, etc.). In some embodiments, a subject is a male or female of any age (e.g., a man, a woman, or a child).

[0081] As used herein, the term “autism spectrum disorder” refers to a range of neurodevelopmental disorders associated with impairments in social interaction, developmental language and communication skills, and repetitive behaviors. For example, the standardized criteria for diagnosing autism spectrum disorder established by the Centers for Disease Control and Prevention (CDC) include 1) persistent deficits in social communication and social interaction and 2) restricted, repetitive patterns of behavior, interests, or activities. Autism spectrum disorders include, for example, autism disorder (also known as “classical autism”), Asperger’s Syndrome, and pervasive developmental disorder (also known as “atypical” autism).

[0082] As used herein, the term "recursive quantitative analysis" ("RQA") refers to the nonlinear data analysis of the quantity and duration of recursions in a dynamic system. RQA is used to characterize the behavior of a dynamic system in phase space.

[0083] As used in this paper, the term "recursion graph" refers to the graphical visualization of time-dependent periodic structures in experimental data.

[0084] As used in this paper, the term "trace" refers to the time-dependent abundance (or concentration) of an elemental isotope. A trace contains multiple data points, each of which is associated with both a time-dependent and an abundance-dependent measure.

[0085] As used herein, the term "feature" refers to, for example, a dynamic periodicity feature extracted from a time-dependent abundance trace of an elemental isotope or a combination of two or more time-dependent abundance traces of an elemental isotope using RQA.

[0086] As used herein, the term “mean diagonal length” (“MDL”) refers to a key metric derived from RQA that reflects a simple measurement of the average length of the diagonals present in a two-dimensional recurrence graph. This metric can serve as an absolute indicator of the duration of periodic components in a given signal.

[0087] As used in this paper, the term "determinism," related to the average diagonal length, refers to the relative ratio of periodic to non-periodic components in a recursive analysis. Determinism indicates the overall periodicity of a given signal.

[0088] As used in this paper, the term “recursion time” (“RT2”) refers to the average time interval between diagonal elements, i.e., the interval between periods.

[0089] As used in this paper, the term "entropy" refers to the variability of the distribution of the average diagonal length, where low-entropy signals exhibit small complexity in the distribution of periodic components, and high-entropy signals exhibit diversity in both short-duration and long-duration periodicity.

[0090] As used herein, the term “capture time” (“TT”) refers to the average length of a hierarchical (vertical or horizontal) structure in a two-dimensional recursive graph that indicates a steady state, similar to how the average diagonal length captures the duration of a periodic process.

[0091] As used in this paper, the term "hierarchy" refers to an overall measure of signal stability. Hierarchy quantifies the ratio of recursion points belonging to a hierarchical structure to the total frequency of recursion points.

[0092] The terminology used herein is for descriptive purposes only and is not intended to be limiting. As used herein, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well. Furthermore, where the terms “including / include,” “having / has / with,” or variations thereof are used in the detailed description and / or claims, such terms are intended to be inclusive in a manner similar to the term “comprising.”

[0093] The following examples illustrate several aspects. It should be understood that many specific details, relationships, and methods are described to provide a complete understanding of the features described herein. However, those skilled in the art will readily recognize that the features described herein can be practiced without one or more of the specific details stated herein, or in other ways. The features described herein are not limited to the described order of actions or events, as some actions may occur in a different order and / or simultaneously with other actions or events. Furthermore, implementing the methods according to the features described herein does not require all the described actions or events.

[0094] Detailed reference will now be made to the embodiments, examples of which are illustrated in the accompanying drawings. Numerous specific details are set forth in the following detailed description to provide a thorough understanding of this disclosure. However, it will be apparent to those skilled in the art that this disclosure can be practiced without these specific details. In other instances, well-known methods, procedures, components, circuits, and networks have not been described in detail so as not to unnecessarily obscure various aspects of the embodiments.

[0095] Example system implementation.

[0096] Having now provided an overview of some aspects of this disclosure, details of an exemplary system will now be described in conjunction with Figure 1. Figure 1AA block diagram of an example computing device 100 according to some embodiments of the present disclosure is shown. In some embodiments, device 100 includes one or more processing units, CPU 102 (also referred to as a processor), one or more network interfaces 104, user interface 106, non-persistent memory 111, persistent memory 112, and one or more communication buses 114 for interconnecting these components. The one or more communication buses 114 optionally include circuitry (sometimes referred to as a chipset) that interconnects and controls communication between system components. Non-persistent memory 111 typically includes high-speed random access memory such as DRAM, SRAM, DDR RAM, ROM, EEPROM, flash memory, while persistent memory 112 typically includes CD-ROM, digital versatile disc (DVD) or other optical storage devices, magnetic tape cassettes, magnetic tape, disk storage devices or other magnetic storage devices, disk storage devices, optical disc storage devices, flash memory devices or other non-volatile solid-state storage devices. Persistent memory 112 optionally includes one or more storage devices located remotely from CPU 102. The one or more non-volatile storage devices within persistent memory 112 and non-persistent memory 111 include non-transitory computer-readable storage media. In some implementations, non-permanent memory 111 or (alternatively) the non-transitory computer-readable storage medium (sometimes in combination with permanent memory 112) stores programs, modules, and data structures or subsets thereof:

[0097] • An optional operating system 116, which includes programs for handling various basic system services and for performing hardware-related tasks;

[0098] • Optional network communication module (or instruction) 118, which is used to connect system 100 to other devices and / or to communication network 104;

[0099] • Optional classifier training module 120, which is used to train a classifier for assessing biological symptoms related to metal metabolism in subjects;

[0100] • Optional data storage, the data storage being used for a dataset of biological samples from training subjects 122, the dataset containing feature data of one or more training subjects 124, wherein the feature data contains parameters relating to each of the features 126 and a diagnostic status 128 (e.g., an indication that the corresponding training subject has been diagnosed with a biological condition related to metal metabolism or has not yet been diagnosed with a biological condition related to metal metabolism).

[0101] • Optional classifier verification module 130, which is used to verify a classifier that distinguishes biological symptoms related to metal metabolism;

[0102] • Optional data storage, said data storage being used for a dataset from biological samples of validation subjects 132; and

[0103] • Optional patient classification module 134, for example, as trained using classifier training module 120, is used to classify subjects as having biological conditions related to metal metabolism.

[0104] In various embodiments, one or more of the aforementioned elements are stored in one or more of the previously mentioned storage devices and correspond to sets of instructions for performing the functions described above. The aforementioned modules, data, or programs (e.g., instruction sets) do not need to be implemented as separate software programs, processes, datasets, or modules, and therefore, subsets of these modules and data can be combined or otherwise rearranged in various embodiments. In some embodiments, non-persistent memory 111 optionally stores a subset of the aforementioned modules and data structures. Furthermore, in some embodiments, the memory stores additional modules and data structures not described above. In some embodiments, one or more of the aforementioned elements are stored in a computer system outside the computer system of the visualization system 100, which is addressable by the visualization system 100, such that the visualization system 100 can retrieve all or part of such data when needed.

[0105] In some embodiments, system 100 is connected to or includes one or more analytical devices for performing chemical analysis. For example, an optional network communication module (or command) 118 is configured to connect system 100 to the one or more analytical devices, for example, via communication network 104. In some embodiments, the one or more analytical devices include a laser ablation inductively coupled plasma mass spectrometer (LA-ICP-MS).

[0106] Although Figure 1 depicts "System 100", the figure is intended more as a functional description of the various features that may exist in a computer system than as a structural schematic diagram of the implementation described herein. In practice, and as those skilled in the art will recognize, items shown individually may be combined, and some items may be separated. Furthermore, although Figure 1 depicts some data and modules in non-persistent memory 111, some or all of these data and modules may reside in persistent memory 112.

[0107] Classification methods.

[0108] Although a system according to this disclosure has been disclosed with reference to Figure 1, in conjunction with Figures 2A-2FDetailed procedures and features of a method 200 for assessing a subject’s metal metabolism-related biological condition based on a biological sample, according to this disclosure, are provided.

[0109] As defined above, metal metabolism-related biological samples (also referred to herein as "biological samples") comprise deposits of certain metals and are growth-related human biological samples (e.g., hair, nails, and teeth). The metal metabolism-related biological samples of this disclosure require expression of growth along a reference line such that the abundance of certain metal deposits can be detected relative to time. In some embodiments, the metal metabolism-related biological sample comprises a hair shaft, wherein the reference line corresponds to a line along the longitudinal direction of the hair shaft. In some embodiments, the metal metabolism-related biological sample comprises a tooth, wherein the reference line corresponds to a nascent line on the enamel surface of the tooth. In some embodiments, the metal metabolism-related biological sample comprises a nail, wherein the reference line corresponds to a line in the nail growth direction. For example, the reference line extends from the nail root toward the nail tip.

[0110] In some embodiments, method 200 includes obtaining a biological sample (e.g., a strand of hair containing a hair shaft) (202). The subject is a human being. In some embodiments, the subject is a child aged 5 years or less (e.g., a child aged 5 years, 4 years, 3 years, 2 years, 1 year, 9 months, 6 months, 3 months, or 1 month). In some embodiments, the subject is an adult. Figure 2B Part I provides exemplary images of hair samples containing hair shafts from a subject according to some embodiments of this disclosure. The hair samples can be simply cut from the subject (e.g., with the aid of scissors). Therefore, the method for obtaining hair samples is non-invasive. The minimum length of the obtained hair sample is 1 cm (e.g., the length of the hair sample is 1 cm, 2 cm, 3 cm, 4 cm, or 5 cm). The hair sample can contain any portion of the hair (e.g., the tip or the portion between the tip and the hair follicle). Specifically, there is no particular requirement that the hair sample contain a hair follicle. Figure 2B Part II provides exemplary images of tooth samples of subjects according to some embodiments of this disclosure. Figure 2B Section III provides exemplary images of nail samples from a subject according to some embodiments of this disclosure. In the case of teeth or hair, obtaining a biological sample means positioning the subject so that samples can be taken from the teeth or nails.

[0111] In some embodiments, the obtained biological sample is pretreated (204) by washing the biological sample with one or more solvents and / or surfactants and drying it. In the case where the biological sample is hair, the hair sample is subjected to TRITON... and ultrapure metal-free water (e.g., The sample is washed in water and dried overnight in an oven (e.g., at 60°C). Pretreatment further comprises preparing a capillary for measurement by placing it on a slide (e.g., a microscopic slide) with an adhesive film (e.g., double-sided tape). The capillary is positioned such that it is substantially straight. The slide with the capillary is then placed in a laser ablation inductively coupled plasma mass spectrometer (LA-ICP-MS) for analysis (206). In the case of biological samples that are teeth or nails, the surface of the biological sample is cleaned (e.g., with a surfactant, water, or one or more solvents). The subject is positioned near the LA-ICP-MS for analysis.

[0112] In some embodiments, LA-ICP-MS analysis includes pre-ablation of the biological sample to remove surface debris and / or impurities. Pre-ablation is performed using such low laser energies, such that it releases particles only on the surface of the biological sample, without releasing particles from beneath the surface. For example, a laser wavelength of 193 nm and less than 0.4 J / cm² are used. 2 The laser energy (e.g., 0.4 J / cm²) 2 0.3J / cm 2 0.2J / cm 2 Or 0.1 J / cm 2 Pre-ablation is performed. In some embodiments, the laser energy is at 0.2 J / cm². 2 Up to 0.4 J / cm 2 Within the range.

[0113] Following pre-ablation, method 200 includes sampling the biological sample with a laser to obtain an ion sample (208) from a corresponding location along a reference line of the biological sample. As explained above, in the case of the capillary, the reference line corresponds to a line along the longitudinal direction of the capillary. For example, Figure 2B Part I shows the hair shaft and a reference line 201 along the longitudinal direction of the hair shaft. In the case of teeth, the reference line corresponds to the nascent line on the enamel surface of the tooth. For example, Figure 2B Part II shows a tooth 220 comprising a portion of enamel 226 and primary dentin 224. Reference line 222 corresponds to the necrophyseal line of tooth 220. In this text, a necrophyseal line refers to a specific incremental growth band on the enamel portion of a tooth. In the case of a nail, the reference line corresponds to a line in the direction of nail growth. For example, Figure 2B Part III shows a fingernail 230 and a reference line 232 extending from the nail root towards the nail tip. Sampling involves irradiating the biological sample with a laser beam (e.g., laser ablation of the hair shaft) and ionizing the plurality of particles using inductively coupled plasma mass spectrometry. For example, Figure 2BRegions 200A and 200B in Part I correspond to exemplary locations along the hair shaft that were irradiated with a laser during laser ablation. A mass spectrometer was used to analyze the ion sample (210) obtained from each corresponding location. Figure 2C Exemplary schematic diagrams of laser sampling of hair shafts of a subject according to some embodiments of the present disclosure are provided. Figure 2C A laser 202 irradiates region 200C on the hair shaft, thereby releasing particles 204. Particles 204 are ionized by inductively coupled plasma (ICP) and further analyzed by mass spectrometry (MS).

[0114] In some embodiments, a wavelength of 193 nm and a laser energy between 0.6 and 1.5 J / cm² are used. 2 Within the range (e.g., laser energy of 0.6 J / cm²). 2 0.7J / cm 2 0.8J / cm 2 0.9J / cm 2 1.0 J / cm 2 1.1 J / cm 2 1.2J / cm 2 1.3J / cm 2 1.4J / cm 2 Or 1.5J / cm 2 The laser is used to perform laser irradiation. In some embodiments, the laser energy is between 0.9 and 1.3 J / cm². 2 Within the range of [specific parameters]. In some embodiments, the laser beam diameter is in the range of 25 micrometers to 35 micrometers (e.g., 25, 27.5, 30, 32.5, or 35 micrometers). In some embodiments, the laser beam diameter is 30 micrometers. When sampling the capillary, the laser beam size, wavelength, and / or laser energy are adjusted such that the laser sampling ablates most of the capillary without releasing any particles from the adhesive film and / or the glass slide holding the capillary.

[0115] Repeated laser irradiation, and at multiple locations along the biological sample (e.g., Figure 2B Elemental isotope data are collected sequentially in regions 200A and 200B of the hair shaft in part I. In some embodiments, the plurality of locations along a reference line of the biological sample comprises at least 100 locations (e.g., 100, 150, 200, 250, 300, 350, 400, 450, or 500 locations). In some embodiments, the corresponding locations (e.g., Figure 2B Regions 200A and 200B in part I are adjacent to each other. Through this method, each region (e.g., regions 200A and 200B) corresponding to different locations on the biological sample is thus correlated with elemental isotopes (e.g., Figure 2CThe abundance of the metal isotopes Zn, Fe, Pb, and Mn is shown in the diagram. In some embodiments, the corresponding locations are spaced at predetermined distances. In some embodiments, sampling is performed along a reference line of the biological sample, starting from the corresponding location closest to the hair tip (e.g., at the location corresponding to the youngest subject). Typically, sampling can begin from the corresponding location closest to the tip or root, provided the sampling direction is known and analysis is performed using an appropriate trained classifier.

[0116] Laser sampling thus generates a set of data points. Each set of data points corresponds to the abundance (e.g., concentration) of a corresponding elemental isotope measured at multiple locations along the biological sample. Each location on a reference line of the biological sample corresponds to a specific growth time of the biological sample. In some embodiments, in the case of hair shaft, each location corresponds to a hair growth period of approximately 130 minutes (e.g., a hair growth period calculated using a 30-micron laser beam size and an average hair growth rate of 1 cm per month). By correlating the multiple locations along the reference line of the biological sample with the corresponding growth time periods, a first dataset comprising multiple traces is obtained. Each trace contains the time-dependent abundance of a corresponding elemental isotope measured from the biological sample.

[0117] Figure 2D Exemplary illustrations of traces 208 according to some embodiments of the present disclosure are provided. Figure 2D Each data point corresponds to the abundance (i.e., count ratio on the y-axis) of a specific elemental isotope measured at multiple locations along the biological sample (i.e., laser distance on the bottom x-axis). The distance the laser travels along the biological sample corresponds to the estimated growth (i.e., biological time) of the biological sample, as shown on the top x-axis. For example, Figure 2D The abundance of specific elemental isotopes in hair measured along a distance of 1.2 cm (12,000 micrometers) is shown. This distance corresponds to a biological time of approximately 35 days. The biological time was estimated using the average hair growth rate (e.g., 1 cm per month).

[0118] In some embodiments, the plurality of element isotopes are selected from the element isotopes listed in Table 1. In some embodiments, the plurality of element isotopes comprises at least 50%, 60%, 70%, 80%, or 90% of the isotopes included in Table 1.

[0119] Table 1: List of elemental isotopes

[0120]

[0121]

[0122] In some embodiments, method 200 includes analyzing (212) a first dataset comprising a plurality of obtained traces, wherein each trace corresponds to the time-dependent abundance (e.g., time-dependent concentration) of a corresponding elemental isotope. In some embodiments, the data analysis includes performing custom operations to clean the data (214). In some embodiments, data cleaning includes smoothing the data over a time span and / or removing data points that are above or below a predetermined threshold. In some embodiments, the data analysis includes removing data points from the traces whose mean absolute difference between adjacent data points is three times the standard deviation of the mean absolute difference between adjacent data points. Figure 2D The operation of removing data points above a predetermined threshold is demonstrated. Peak 210 corresponds to a data point where the mean absolute difference between adjacent data points exceeds three times the standard deviation of the mean absolute difference between adjacent data points. Therefore, peak 210 is removed from trace 208.

[0123] In some embodiments, the analytical dataset further includes normalization for each trace against an internal standard. In some embodiments, when the sample is a hair shaft, the internal standard is sulfur, which is the most abundant elemental isotope in hair and can therefore be used as a measure of hair density and / or stiffness. However, in practice, any element detected in a sample uniformly incorporated during the development / growth of a biological sample that does not fluctuate with environmental exposures (e.g., diet) can be used as an internal standard, including any element disclosed in the tables of this disclosure. For example, in the case of a tooth sample, bismuth-209 can be used as an internal standard.

[0124] Method 200 includes performing recursive quantitative analysis (RQA) to analyze a first dataset containing time-dependent traces of elemental isotopes to obtain a set of features describing the dynamic periodicity of the traces. RQA measures the variability in the time-dependent traces of elemental isotopes. RQA involves estimating features describing the periodicity of a given waveform, including determinism, average diagonal length, and entropy. The methods and features of RQA are described, for example, in Webber et al., “Simpler Methods Do It Better: Success of Recurrence Quantification Analysis as a General Purpose Data Analysis Tool,” *Physics Letters A* 373, 3753-3756 (2009) and Marwan et al., “Recurrence Plots for the Analysis of Complex Systems,” *Physics Reports* 438, 237-239 (2007), the contents of each of which are incorporated herein by reference in their entirety. In some embodiments, the time-dependent traces of elemental isotopes are analyzed using other analytical methods known in the art, such as Fourier transform, wavelet analysis, and cosine analysis. Such methods can be applied to derive similar metrics, including spectral analysis of frequency components and their associated powers. For predictive classification purposes, these measures and related derived metrics can be used to replace features derived from RQA to analyze time-dependent traces of elemental isotopes obtained from biological samples.

[0125] RQA includes constructing a recursive graph (216) that visualizes and analyzes the dynamic temporal structure in the corresponding obtained traces. Figure 2E Exemplary illustrations are provided of variations in the abundance of individual isotopes derived from corresponding traces according to some embodiments of this disclosure. Figure 2E Part I shows the traces corresponding to the time-dependent abundance (or concentration) of copper (Cu) as measured from the subject's hair shaft. The y-axis shows the measured abundance of copper, and the x-axis shows the sequential measurements along the hair shaft, reflecting the longitudinal increment over time. Figure 2EPart II presents the phase map derived from the traces in Part I. Based on the one-dimensional traces measured from the capillary, an additional dimension is calculated to embed the traces in a higher-dimensional space called the phase map, where t refers to the value of the original trace, and the dimensions (t+τ) and (t+2τ) are derived by lags the original time series by an interval τ. The embedded phase map is then subjected to further analysis to construct a recurrence graph and recurrence quantitative analysis (RQA). Part III shows the recurrence quantitative map of copper isotopes derived from the phase map presented in Part II. The RQA method examines the delay intervals between states in a given system, where black dots reflect the time intervals when the system revisits the same state. Periodic processes where the system continuously repeats a given state pattern will be represented by diagonal black lines in the recurrence graph, while stable periods will be represented by square structures, pseudo-repetitions by black dots, and unique events by blank spaces.

[0126] In some embodiments, the recurrence graph is constructed for traces of a single elemental isotope or a combination of two elemental isotopes (e.g., for elemental isotopes selected from Table 1). For example, Figure 2E A recurrence relation for copper isotopes is shown. Alternatively, a recurrence relation is constructed to visualize the interaction periodic patterns of isotopes of two elements. In some embodiments, the recurrence relation is constructed for combinations of three or more element isotopes.

[0127] Method 200 further includes analyzing the recurrence graph to obtain a set of features associated with the recurrence graph (218). Features, which may be interchangeably referred to as “rhythmic features” or “dynamic features,” provide quantitative measures describing the periodicity present in the plurality of traces. The features are selected from mean diagonal length (MDL), determinism (or predictability), recurrence time (RT), entropy, capture time (TT), and hierarchicality. The definition of each of these feature types is provided above in the definition section.

[0128] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 2.

[0129] In some embodiments, the set of features includes all the features listed in Table 2.

[0130] In some embodiments, the set of features comprises at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 2. In some embodiments, features derived from Table 2 in this manner, according to this disclosure, are considered “core” features for assessing a subject’s primary biological symptom (e.g., autism spectrum disorder, etc.). In some embodiments, the set of features further comprises one or more features listed in Table 3 (in addition to the core features).

[0131] Table 2: A list of characteristics associated with their corresponding elemental isotopes or corresponding combinations of two elemental isotopes.

[0132] Determinism (Determinism_Cd) Cd Determinism (Determinism_Cr) Cr Determinism (Determinism_ZnHg) ZnHg Determinism (Determinism_Cu) Cu Determinism (Determinism_ZnMn) ZnMn Determinism (Determinism_Sr) Sr Entropy (Entropy_As) As Determinism (Determinism_Mg) Mg Entropy (Entropy_Li) Li Determinism (Determinism_ZnCu) ZnCu Entropy (Entropy_ZnCu) ZnCu Average diagonal length (MDL_ZnCu) ZnCu Determinism (Determinism_Ca) Ca Determinism (Determinism_Mn) Mn Determinism (Determinism_Ni) Ni Determinism (Determinism_ZnMg) ZnMg Determinism (Determinism_ZnCr) ZnCr Determinism (Determinism_Pb) Pb Determinism (Determinism_ZnNi) ZnNi Determinism (Determinism_ZnSn) ZnSn Determinism (Determinism_Li) Li Determinism (determinism_Hg) Hg Determinism (Determinism_Fe) Fe Determinism (Determinism_As) As Determinism (Determinism_ZnI) ZnI

[0133] Table 3: A list of additional characteristics associated with their corresponding elemental isotopes or corresponding combinations of two elemental isotopes.

[0134]

[0135]

[0136]

[0137]

[0138]

[0139]

[0140] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 3. In some embodiments, the set of features includes all the features listed in Table 3. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 3.

[0141] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Tables 2 and 3. In some embodiments, the set of features includes all the features listed in Tables 2 and 3. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Tables 2 and 3.

[0142] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 4. In some embodiments, the set of features includes all the features listed in Table 4. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 4.

[0143] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 5. In some embodiments, the set of features includes all the features listed in Table 5. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 5.

[0144] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 6. In some embodiments, the set of features includes all the features listed in Table 6. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 6.

[0145] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 7. In some embodiments, the set of features includes all the features listed in Table 7. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 7.

[0146] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 8. In some embodiments, the set of features includes all the features listed in Table 8. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 8.

[0147] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 9. In some embodiments, the set of features includes all the features listed in Table 9. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 9.

[0148] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in Table 10. In some embodiments, the set of features includes all the features listed in Table 10. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 10.

[0149] In some embodiments, the set of features associated with each feature and a corresponding elemental isotope or combination of elemental isotopes (e.g., a combination of two elemental isotopes, or a combination of more than two elemental isotopes) is selected from the features listed in any combination of Tables 2, 3, 4, 5, 6, 7, 8, 9, and 10. In some embodiments, the set of features includes all the features listed in Tables 2, 3, 4, 5, 6, 7, 8, 9, and 10. In some embodiments, the set of features includes at least 5%, 10%, 15%, 20%, or 25% of the features listed in Tables 2, 3, 4, 5, 6, 7, 8, 9, and 10.

[0150] Method 200 further includes inputting the obtained set of features into a trained classifier (220). In some embodiments, the trained classifier includes a predictive computation algorithm (222) for obtaining the probability that a subject has a biological symptom related to metal metabolism. In some embodiments, the predictive computation algorithm calculates Equation 1:

[0151]

[0152] in

[0153] p(subject) is the probability that the subject has the aforementioned biological condition related to metal metabolism.

[0154] e is the Euler number.

[0155] α is when β1x1+…+β k x k When equal to zero, it is a calculated parameter related to the probability that the subject suffers from the aforementioned biological symptom associated with metal metabolism.

[0156] β 1,…,k The weight parameters correspond to the weights associated with each feature in the set of features from features 1 to k, and

[0157] x 1,…,k The value corresponding to each feature in the set of features, which includes features 1 to k.

[0158] Features 1 to k are selected from the features listed in Table 2, and optionally additionally from Table 3. Weight parameter β 1,…,k It is defined based on classifier training. The probability p (subject) is set to a number in the range of 0 to 1, where 1 corresponds to a 100% probability that the subject has a biological symptom related to metal metabolism.

[0159] In some embodiments, method 200 further includes applying a predetermined threshold to the obtained probability p(subject) (224). If the obtained probability p(subject) is higher than the predetermined threshold, the subject is assessed as having a biological condition related to metal metabolism. If the obtained probability is lower than the predetermined threshold, the subject is assessed as not having a biological condition related to metal metabolism. In some embodiments, the predetermined threshold is between 0.3 and 0.6 (e.g., the predetermined threshold is 0.3, 0.35, 0.4, 0.45, 0.5, 0.55, or 0.6). In some embodiments, the predetermined threshold is 0.45. In some embodiments, the obtained probability is represented by an associated odds (e.g., an odds ratio (OR), which can be derived from the probability such that OR = p / (1-p)). For example, the assessment includes assessing the odds that the subject has a biological condition related to metal metabolism.

[0160] In some embodiments, method 200 further includes distinguishing between a first biological symptom related to metal metabolism and a substitute symptom, such as a second biological symptom related to metal metabolism. In some embodiments, the substitute symptom is associated with an unknown symptom (e.g., neurotypical symptom (NT)). In some embodiments, the first biological symptom related to metal metabolism is associated with autism spectrum disorder (ASD), and the substitute symptom is associated with attention deficit / hyperactivity disorder (ADHD). In some embodiments, the substitute symptom is any other neurodevelopmental symptom, or a comorbid diagnosis of two neurodevelopmental symptom. Figure 2F Illustrations are provided of experimental data distinguishing autism spectrum disorder (ASD) from other neurodevelopmental disorders according to descriptions of some embodiments of this disclosure. It should be noted that, based on... Figure 2F Based on the experimental data shown, the method 200 of this disclosure can distinguish between autism spectrum disorder and ADHD. As shown, this disclosure can also distinguish between autism spectrum disorder and comorbid (CM) cases diagnosed with both autism spectrum disorder and ADHD.

[0161] The process and characteristics of a method 200 for assessing metal metabolism-related biological symptoms in subjects based on biological samples have now been disclosed with reference to Figure 2. Figures 3A-3ETogether, flowcharts are provided of the basic procedures and characteristics of a method 3000 for assessing a subject's metal metabolism-related biological sample according to some embodiments of this disclosure, wherein optional boxes are indicated by dashed boxes. In some embodiments, method 3000 corresponds to method 200.

[0162] Figure 3A Box 3100. Method 3000 includes, for example, sampling each of a plurality of locations at a reference line along a metal metabolism-related biological sample of the subject using a laser (e.g., LA-ICP-MS), thereby obtaining a plurality of ion samples (e.g., Figure 2B Regions 200A and 200B of the hair shaft in Part I). Each of the plurality of ion samples corresponds to a different position among the plurality of locations, and each of the plurality of locations represents a different growth stage of the biosample associated with metal metabolism.

[0163] Figure 3A Box 3200. Method 3000 includes analyzing each of the plurality of ion samples using a mass spectrometer, thereby obtaining a first dataset. The first dataset contains multiple traces (e.g., Figure 2D (trace 208 in the sample). Each of the plurality of traces represents the concentration of a corresponding elemental isotope among multiple elemental isotopes determined over time based on the concentration of the corresponding elemental isotopes from the plurality of ion samples.

[0164] Figure 3A Box 3300. Method 3000 includes deriving a second dataset from the plurality of traces containing a set of features (e.g., a set of features selected from the features listed in Table 2). Each corresponding feature in the set of features is determined by variations in a single isotope or combination of isotopes in the plurality of traces. For example, Figure 2E Part III shows from Figure 2E The recurrence plot of copper isotopes was derived from the traces of part II. The variation in copper isotope abundance was observed to follow a diagonal pattern in the recurrence plot.

[0165] Figure 3A Box 3400. In some embodiments, method 3000 further includes inputting the set of features into a trained classifier, thereby obtaining from the trained classifier the probability that the subject has the first biological symptom related to metal metabolism. In some embodiments, the trained classifier is a neural network algorithm, a support vector machine algorithm, a decision tree algorithm, an unsupervised clustering model algorithm, a supervised clustering model algorithm, or a regression model.

[0166] Figure 3BBox 3110. In some embodiments, sampling the hair shaft includes irradiating the subject's metal metabolism-related biological sample with the laser, thereby extracting a plurality of particles from the subject's metal metabolism-related biological sample, and ionizing the plurality of particles using an inductively coupled plasma mass spectrometer, thereby obtaining the plurality of ion samples (e.g., Figure 2C ).

[0167] Figure 3B Box 3120. In some embodiments, along the plurality of locations of the hair shaft (e.g., Figure 2B Regions 200A and 200B of the hair shaft in part I are arranged in order such that a first position along the plurality of locations of the subject’s metal metabolism-related biological sample corresponds to the position closest to the tip of the subject’s metal metabolism-related biological sample.

[0168] Figure 3B Box 3130. Method 3000 further includes pretreating the subject's metal metabolism-related biological sample with a solvent or surfactant prior to sampling the hair shaft of the subject. For example, pretreating the hair shaft with TRITON... and ultrapure metal-free water (e.g., Wash with water and dry overnight in an oven (e.g., at 60 degrees Celsius).

[0169] Figure 3B Box 3140. Method 3000 further includes irradiating the subject's metal metabolism-related biological sample with a low-power laser before sampling the hair shaft of the subject to remove any debris (e.g., pre-ablation of hair shaft, teeth, or nails) from the subject's metal metabolism-related biological sample. For example, a laser wavelength of 193 nm and less than 0.4 J / cm² are used. 2 The laser energy (e.g., 0.4 J / cm²) 2 0.3J / cm 2 0.2J / cm 2 Or 0.1 J / cm 2 Pre-ablation is performed. In some embodiments, the laser energy is at 0.2 J / cm². 2 Up to 0.4 J / cm 2 Within the range.

[0170] Figure 3B Box 3141. The biosamples of the subjects related to metal metabolism were selected from a group consisting of hair shafts, teeth, and nails (e.g., Figure 2B (The hair shaft, teeth, and nails are shown in Parts I, II, and III, respectively).

[0171] Figure 3B Box 3141-1. The biosample related to metal metabolism of the subject is the hair shaft, and the reference line corresponds to the longitudinal direction of the hair shaft (e.g., Figure 2B Reference line 201 in Part I).

[0172] Figure 3B Box 3141-1. The metal metabolism-related biological sample of the subject is the tooth, and the reference line corresponds to the neoplasm on the enamel surface of the tooth (e.g., along...). Figure 2B Reference line 222 for the neonatal line of tooth 220 in Part II). In some embodiments, the metal metabolism-related biological sample of the subject is a fingernail, and the reference line corresponds to a line extending from the root of the fingernail to the tip of the fingernail (e.g., Figure 2B Reference line 232 for nail 230 in Part III).

[0173] Figure 3C Box 3210. The plurality of element isotopes are selected from the element isotopes listed in Table 1. In some embodiments, the plurality of element isotopes comprises at least 50%, 60%, 70%, 80%, or 90% of the isotopes included in Table 1.

[0174] Figure 3C Box 3220. Each of the plurality of traces contains a plurality of data points. Each data point is an instance of the corresponding location among the plurality of locations. In some embodiments, each trace contains at least 100 locations (e.g., 100, 150, 200, 250, 300, 350, 400, 450, or 500 locations). In some embodiments, each data point corresponds to a hair growth period of approximately 130 minutes (e.g., a hair growth period calculated using a 30-micron laser beam size and an average hair growth rate of 1 cm per month).

[0175] Figure 3C Box 3230. The concentration of the corresponding element isotope corresponds to the relative abundance of the corresponding element isotope relative to the control element isotope. The control element isotope is contained in the plurality of ion samples. In some embodiments, the control element isotope is sulfur.

[0176] Figure 3DBox 3310. The set of features is selected from the features listed in Table 2. In some embodiments, the set of features includes the features listed in Table 2. In some embodiments, the set of features includes at least 50%, 60%, 70%, 80%, or 90% of the features listed in Table 2. Each feature in the set of features is associated with a single corresponding trace or two corresponding traces of the plurality of traces.

[0177] Figure 3D Box 3320. In addition to the features selected from those listed in Table 2, the set of features further includes one or more features listed in Table 3.

[0178] Figure 3D Box 3330. The process of deriving the second dataset involves removing such data points from the plurality of data points that do not meet the first criterion. In some embodiments, the first criterion includes the mean absolute difference between adjacent data points in the plurality of data points being three times the standard deviation of the mean absolute difference between adjacent data points (e.g., from...). Figure 2D Peak 210 was removed from trace 208.

[0179] Figure 3D Box 3340. The set of features is selected from average diagonal length, determinism, recursion time, entropy, capture time, and hierarchicality.

[0180] Figure 3E Box 3410. In some embodiments, the trained classifier calculates:

[0181]

[0182] Where p (subject) is the probability that the subject suffers from the biological condition related to metal metabolism, e is the Euler number, and α is the probability that β1x1 + ... + β k x k When equal to zero, β is a calculated parameter associated with the probability that the subject suffers from the aforementioned biological symptom related to metal metabolism. 1,…,k The weight parameters correspond to the weights associated with each of the set of features from features 1 to k, and x 1,…,k The value derived corresponding to each feature in the set of features, which includes features 1 to k.

[0183] Figure 3E Box 3420. Based on the determination that p (subject) is higher than a predetermined threshold, the subject is confirmed to have the aforementioned biological condition related to metal metabolism.

[0184] Figure 3EBox 3500. In some embodiments, assessing the subject's metal metabolism-related biological condition further includes distinguishing between a first metal metabolism-related biological condition and a second metal metabolism-related biological condition, the second biological condition being different from the first metal metabolism-related biological condition.

[0185] Figure 3E Box 3510. In some embodiments, the first biological symptom is autism spectrum disorder, and the second biological symptom is attention deficit / hyperactivity disorder.

[0186] Figure 3E Box 3512. In some embodiments, the first biological symptom associated with metal metabolism is selected from the group consisting of: autism spectrum disorder (ADS), attention deficit / hyperactivity disorder (ADHD), amyotrophic lateral sclerosis (ALS), schizophrenia, irritable bowel disease (IBD), pediatric kidney transplant rejection, and pediatric cancer.

[0187] In some embodiments, regarding Figures 3A-3E The described method 3000 is performed by means of an apparatus that executes one or more programs (e.g., one or more programs stored in non-persistent memory 111 in FIG. 1 or stored in persistent memory 112), said one or more programs containing instructions for executing method 3000. In some embodiments, method 3000 is executed by a system including at least one processor (e.g., processing core 102) and memory (e.g., one or more programs stored in non-persistent memory 111 or stored in persistent memory 112), said memory including instructions for executing method 3000.

[0188] Classifier training.

[0189] Now it has been referenced Figures 3A-3E The method and features of method 3000 are disclosed. Figure 4A flowchart of the process and features of a method 4000 for training a classifier for assessing metal metabolism-related biological conditions in subjects, according to some embodiments of this disclosure, is provided, wherein optional boxes are indicated by dashed boxes. The method for training the classifier includes collecting metal metabolism-related biological samples from a plurality of training subjects, and training the classifier using the collected biological samples. The training subjects are humans. Each training session has a diagnostic status indicating whether the training subject has been diagnosed with a metal metabolism-related biological condition or has not yet been diagnosed with a metal metabolism-related biological condition. In some embodiments, the training subjects are children aged 5 years or less (e.g., 5 years or less, 4 years, 3 years, 2 years, 1 year, 9 months, 6 months, 3 months, or 1 month). The steps of the method 4000 described below with respect to boxes 4100-4300 are performed for each of the plurality of training subjects.

[0190] Figure 4 Box 4100. Method 4000 includes sampling each of a plurality of corresponding locations on a corresponding reference line on a corresponding biological sample related to metal metabolism of the corresponding training subject using a laser, thereby obtaining a plurality of corresponding ion samples. Each of the plurality of corresponding ion samples corresponds to a different location among the plurality of corresponding locations, and each of the plurality of corresponding locations represents a different growth stage of the corresponding biological sample related to metal metabolism.

[0191] Figure 4 Box 4200. Method 4000 includes analyzing each of the corresponding plurality of ion samples using a mass spectrometer, thereby obtaining a corresponding first dataset containing the corresponding plurality of traces. Each of the corresponding plurality of traces is the concentration of a corresponding element isotope among multiple element isotopes determined over time based on the corresponding plurality of ion samples.

[0192] Figure 4 Box 4300. Method 4000 includes deriving a corresponding second dataset from the corresponding plurality of traces, comprising a corresponding set of features, each corresponding feature in the corresponding set of features being determined by variations in a single isotope or combination of isotopes in the corresponding plurality of traces.

[0193] Figure 4Box 4400. Method 4000 further includes training an untrained or partially untrained classifier using: (i) a corresponding set of features for each corresponding second dataset of each of the plurality of training subjects; and (ii) a corresponding diagnostic state selected from the first diagnostic state and the second diagnostic state for each of the plurality of training subjects. The trained classifier provides an indication of whether the test subject suffers from the first biological symptom related to metal metabolism based on the value of a feature from a set of features obtained from a biological sample related to metal metabolism of the test subject. In some embodiments, (box 4410) the trained classifier is a neural network algorithm, a convolutional neural network algorithm, a support vector machine algorithm, a decision tree algorithm, an unsupervised clustering model algorithm, a supervised clustering model algorithm, or a regression model. In some embodiments, (box 4420) the trained classifier is a multinomial or binomial. In some embodiments, the trained classifier may be used to make a binary prediction about whether a sample originates from a subject with a first biological condition related to metal metabolism; or it may be polynomial, thereby distinguishing undiagnosed subjects from subjects with a first biological condition related to metal metabolism or subjects with a second biological condition related to metal metabolism, wherein the second biological condition is different from the first biological condition.

[0194] In some embodiments, the classifier is a neural network or a convolutional neural network. See Vincent et al., 2010, “Stacked denoising autoencoders: Learning useful representations in a deep network with a local denoising criterion,” *Journal of Machine Learning Research*, 11, pp. 3371-3408; Larochelle et al., 2009, “Exploring strategies for training deep neural networks,” *Journal of Machine Learning Research*, 10, pp. 1-40; and Hassoun, 1995, *Fundamentals of Artificial Neural Networks*, MIT. Each of these references is incorporated herein by reference.

[0195] SVM is described in the following references: Cristianini and Shaw-Taylor, 2000, “An Introduction to Support Vector Machines”, Cambridge University Press, Cambridge; Boser et al., 1992, “A training algorithm for optimal margin classifiers”, Proceedings of the 5th Annual ACM Symposium on Computational Learning Theory, ACM Press, Pa., pp. 142-152; Vapnik, 1998, “Statistical Learning Theory”, Wiley, New York; Mount, 2001, “Bioinformatics: sequence and genome analysis”, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; Duda, “Pattern Classification”, 2nd edition, 2001, John Wiley & Sons. Wiley & Sons, Inc., pp. 259 and 262-265; and Hastie, 2001, *The Elements of Statistical Learning*, Springer, New York; and Furey et al., 2000, *Bioinformatics*, 16, 906-914, each of which is hereby incorporated in full. When used for classification, SVM separates a given set of binary labeled data from a hyperplane that is maximally farthest from the labeled data. For cases where linear separation is not possible, SVM can be operated in conjunction with a “kernel” technique that automatically performs a nonlinear mapping of the feature space. The hyperplane discovered by SVM in the feature space corresponds to the nonlinear decision boundary in the input space.

[0196] Decision trees are generally described in the following literature: Duda, 2001, Pattern Classification, John Wiley & Sons, New York, pp. 395-396, which are hereby incorporated by reference. Tree-based methods divide the feature space into a set of rectangles and then fit a model (e.g., a constant) into each rectangle. In some embodiments, the decision tree is a random forest regression. One particular algorithm that can be used is Classification and Regression Tree (CART). Other specific decision tree algorithms include, but are not limited to, ID3, C4.5, MART, and random forests. CART, ID3, and C4.5 are described in the following literature: Duda, 2001, Pattern Classification, John Wiley & Sons, New York, pp. 396-408 and 411-412, which are hereby incorporated by reference. CART, MART, and C4.5 are described in the following literature: Hastie et al., 2001, Foundations of Statistical Learning, Springer Publishing, New York, Chapter 9, which are hereby incorporated by reference. Random forests are described in the following literature: Breiman, 1999, “Random Forests – Random Features,” Technical Report 567, Department of Statistics, University of California, Berkeley, September 1999, which is hereby incorporated by reference in its entirety.

[0197] Clustering (e.g., unsupervised and supervised clustering algorithms) is described in the following literature: Duda and Hart, *Pattern Classification and Scene Analysis*, pp. 211-256, 1973, John Wiley & Sons, New York (hereinafter referred to as "Duda 1973"), which is incorporated herein by reference in its entirety. As described in Section 6.7 of Duda 1973, the clustering problem is described as the problem of finding natural groupings in a dataset. To determine natural groupings, two problems are solved. First, the method of measuring the similarity (or dissimilarity) between two samples is determined. Using this metric (similarity metric) ensures that samples in one cluster are more similar to each other than samples in other clusters. Second, the mechanism for partitioning the data into clusters using the similarity metric is determined. The similarity metric is discussed in Section 6.7 of Duda 1973, where it is stated that one way to begin a clustering investigation is to define a distance function and compute a matrix of distances between all pairs of samples in the training set. If distance is a good measure of similarity, the distance between reference entities in the same cluster will be significantly smaller than the distance between reference entities in different clusters. However, as described on page 215 of Duda (1973), clustering does not require the use of a distance metric. For example, a non-metric similarity function s(x,x') can be used to compare two vectors x and x'. Typically, s(x,x') is a symmetric function with a larger value when x and x' are "similar" to some extent. An example of the asymmetric similarity function s(x,x') is provided on page 218 of Duda (1973). Once a method for measuring the "similarity" or "dissimilarity" between points in a dataset has been chosen, clustering requires a criterion function that measures the clustering quality of any partitions of the data. Partitions of the dataset that are extreme values ​​of the criterion function are used to cluster the data. See page 217 of Duda (1973). Criterion functions are discussed in Section 6.8 of Duda (1973). Recently, John Wiley & Sons of New York published the second edition of Duda et al.'s *Pattern Classification*. Clustering is described in detail on pages 537-563.More information on clustering techniques can be found in the following literature: Kaufman and Rousseeuw, 1990, *Finding Groups in Data: An Introduction to Cluster Analysis*, Wiley & Co., New York, NY; Everitt, 1993, *Cluster Analysis* (3rd Edition), Wiley & Co., New York, NY; and Backer, 1995, *Computer-Assisted Reasoning in Cluster Analysis*, Prentice Hall, Upper Saddle River, New Jersey, each of which is hereby incorporated by reference. Specific exemplary clustering techniques that may be used in this disclosure include, but are not limited to, hierarchical clustering (merging clustering using nearest neighbor, farthest neighbor, average association, centroid, or sum of squares algorithms), k-means clustering, fuzzy k-means clustering, and Jarvis-Patrick clustering. In some embodiments, clustering including unsupervised clustering is applied, in which there is no preconceived notion of what clusters should be formed when clustering the training set.

[0198] Regression models, such as the multi-category logit model, are described in the following literature: Agresti, An Introduction to Categorical Data Analysis, 1996, John Wiley & Sons, New York, Chapter 8, which is hereby incorporated by reference in its entirety. In some embodiments, the classifier utilizes a regression model disclosed in the following literature: Hastie et al., Foundations of Statistical Learning, 2001, Springer Publishing, New York.

[0199] In some embodiments, regarding Figure 4 The described method 4000 is performed by means of an apparatus that executes one or more programs (e.g., one or more programs stored in non-persistent memory 111 in FIG. 1 or stored in persistent memory 112), said one or more programs containing instructions for executing method 4000. In some embodiments, method 4000 is performed by a system including at least one processor (e.g., processing core 102) and memory (e.g., one or more programs stored in non-persistent memory 111 or stored in persistent memory 112), said memory including instructions for executing method 4000.

[0200] Example.

[0201] Example 1 – Assessing a subject's autism spectrum disorder

[0202] Use about Figures 2A-2F Method 200 described assesses autism spectrum disorder in two subjects (Subject 1 and Subject 2). Table 4 shows the results (e.g., the "Features" column) containing the features (e.g., empirical results, such as x-values) associated with the corresponding parameter estimates β values ​​obtained from the training set, as well as the empirical results for Subject 1 and Subject 2. The β values ​​are obtained by estimating the change in log odds associated with a one-unit change in the corresponding feature for each feature describing autism spectrum disorder status in the training dataset. Given a calculated α parameter of 36.31, for each corresponding subject, the estimated parameter β and x values ​​for each subject are input into the algorithm to calculate p(subject) (see Equation 1 above). For Subject 1, the estimated parameter β and empirical result x yield an estimated probability p(subject 1) of 2.28% that Subject 1 has autism spectrum disorder. For Subject 2, the estimated parameter β and empirical result x yield an estimated probability p(subject 2) of 96.9% that Subject 2 has autism spectrum disorder. Therefore, with a predetermined threshold of 50%, Subject 1 was assessed as not having autism spectrum disorder, and Subject 2 was assessed as having autism spectrum disorder. Furthermore, the probability that Subject 1 had autism spectrum disorder was 0.023, and the probability that Subject 2 had autism spectrum disorder was 31.2. Probabilities were calculated using Equation 2.

[0203]

[0204] Table 4: Features and relevant parameter estimates obtained from the training set, and empirical x-values ​​for Subject 1 and Subject 2.

[0205]

[0206]

[0207] Example 2 – Receiver Operating Characteristic (ROC) Curve.

[0208] Figure 5A Receiver operating characteristic (ROC) curves are shown to illustrate the accuracy of the disclosed methods for assessing autism spectrum disorder in subjects, according to some embodiments. (Regarding...) Figure 5AIn the described experiment, evaluation was performed by measuring the hair shaft of the subjects. ROC curves can be used to evaluate the performance of binary classifiers. ROC curves are plotted as sensitivity (also known as the true positive rate) versus specificity (also known as the true negative rate). A perfect classifier will have 100% sensitivity and 100% specificity, and the area under the curve (AUC) will correspond to 1. Figure 5A As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.947, indicating that the disclosed method has an accuracy of over 90% in assessing whether a subject has autism spectrum disorder.

[0209] Example 3 – Assessing autism spectrum disorder in subjects based on hair samples from one or two parents

[0210] To develop a classifier that could determine whether subjects had autism spectrum disorder, hair samples were collected from the parents (biological mother and father) of twins in a study conducted in Sweden (Roots of Autism and ADHD Study in Sweden – RATSS; Marwan et al., 2007, “Recurrence plots for the analysis of complex systems,” Phys. Rep. 438, 237–329.). The aim of the study was to predict a diagnosis of autism spectrum disorder (ASD) in children solely based on parental hair samples. The children had already undergone clinical testing for autism. No data regarding the children were used in this analysis other than for diagnosis. Three classifiers were developed: a) a classifier that uses only the mother's hair to predict childhood autism (n=29; 14 ASD cases, 15 controls); b) a classifier that uses only the father's hair (n=23; 9 ASD cases and 14 controls); and c) a classifier that uses both the mother's and father's hair (n=52; 23 ASD cases, 29 controls).

[0211] Table 5 shows the features used and their β values ​​for the mother's hair cohort, the father's hair cohort, and the combination of the mother's and father's hair cohorts. The β values ​​were obtained by estimating the change in log odds for each feature in the corresponding cohort that describes the autism spectrum disorder status, associated with a one-unit change in the corresponding feature.

[0212] Figure 5B , 5C5D and 5D respectively demonstrate experimental ROC curves for evaluating the accuracy of trained classifiers for autism spectrum disorder based on mother's hair, father's hair, and combinations of mother's and father's hair, according to some embodiments. Figure 5B As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.886, indicating that the disclosed method has over 85% accuracy in assessing autism spectrum disorder based on hair samples from the subject's mother. Figure 5C As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.800, indicating that the disclosed method has 80% accuracy in assessing autism spectrum disorder based on hair samples from the subject's father. Figure 5D As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.859, indicating that the disclosed method has an accuracy of over 85% for assessing autism spectrum disorder based on a combination of hair samples from the subject's mother and the subject's father.

[0213] Table 5: Characteristics and β values ​​for subjects based on samples obtained from the subject's mother, the subject's father, and combinations of the subject's mother and father.

[0214]

[0215]

[0216]

[0217]

[0218]

[0219] Example 4 – Amyotrophic Lateral Sclerosis (ALS)

[0220] Participants with ALS who met the modified EI Escorial World Federation of Neurology criteria (N=36) were recruited at the ALS outpatient clinic. Clinical and family history data were obtained. Age-matched and sex-matched control participants were recruited at the oral surgery outpatient clinic. Control participants (N=31) or their first- or second-degree family members were excluded if they had a neurodegenerative disease. Informed consent was obtained from participants or their close relatives.

[0221] For ALS, the assessment was performed from tooth samples. Table 6 shows the features used and their corresponding β values. The β values ​​were obtained by estimating the change in the log odds of the corresponding feature describing the ALS state in the respective cohort, which is related to a one-unit change in the corresponding feature. Figure 6 Experimental ROC curves are shown to demonstrate the accuracy of the published method for evaluating ALS across cohorts. Figure 6 As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.869, indicating that the disclosed method has 85% accuracy for cross-cohort assessment of ALS based on tooth samples.

[0222] Table 6: Characteristics and β values ​​of tooth samples from subjects used to assess ALS in subjects.

[0223] (intercept) 83.96232 no Results. Maximum frequency (MaxFreq) of Cu. -1.27115 no Result. Li. Maximum frequency -0.0579 no Result. Mg. Maximum frequency -267.415 no Result. Maximum frequency of Mn. 23.69603 no Result. Maximum frequency of Zn. -35.6081 no Determinism_Cu 48.30136 yes Determinism_Li -96.9188 yes Determinism_Mg 43.43997 yes Determinism_Mn 63.09591 yes Determinism_Zn 123.8952 no Entropy_Cu -68.1936 yes Entropy_Li 61.47467 yes Entropy_Mg -3.63648 yes Entropy_Mn -17.6021 yes Entropy_Zn -28.8004 yes MDL_Cu 8.374507 yes MDL_Li -11.6838 yes MDL_Mg 83.96232 yes MDL_Mn -1.27115 yes MDL_Zn -0.0579 yes

[0224] Example 5 – Schizophrenia

[0225] Participants diagnosed with schizophrenia under the DSM-IV test were selected from the Genetic Risk and Outcomes of Psychosis (GROUP) study (n=20) with unaffected siblings serving as controls (n=7). The severity of positive symptoms, negative symptoms, and general psychopathology was assessed using the Positive and Negative Syndrome Scale (PANSS). Additionally, participants diagnosed with schizophrenia under the DSM-IV test (n=25) and controls (n=24) were selected from the Avon Longitudinal Study of Parents and Children (ALSPAC), a prospective longitudinal cohort study conducted in the UK. The presence of DSM-IV schizophrenia in the ALSPAC was determined at ages 18 and 24 using semi-structured interviews based on the Schedule of Clinical Assessments of Neuropsychiatric Psychiatry (SCAN version 2.0).

[0226] For schizophrenia, assessments were performed using dental samples. Table 7 shows the features used and their corresponding beta values. Beta values ​​were obtained by estimating the change in the log odds of the corresponding feature in each cohort describing the schizophrenia state, relative to a one-unit change in the corresponding feature. Figure 7 The experimental ROC curves for cross-cohort assessment of schizophrenia are shown. Figure 7 As shown, the AUC of the ROC curve corresponds to 1.000, thus indicating that the disclosed method has 100% accuracy in identifying schizophrenia based on cross-cohort dental samples.

[0227] Table 7: Characteristics and β values ​​of dental samples from subjects used to assess schizophrenia.

[0228]

[0229]

[0230] Example 6 – Irritable Bowel Disease (IBD)

[0231] Subjects were recruited from a study conducted in Portugal. Dental samples were obtained from 11 patients diagnosed with IBD (Chron's disease = 6, ulcerative colitis / uncertain colitis = 5) and 16 unaffected controls. All participants were born and raised in the same province of Portugal. IBD was assessed in each subject using a method similar to that described above with respect to Examples 2 and 3. For IBD, the assessment was performed from dental samples. Table 8 shows the features used and their corresponding beta values. Beta values ​​were obtained by estimating the change in the log odds of the characteristic describing IBD status for each feature in the corresponding cohort, with respect to a one-unit change in the corresponding feature.

[0232] Figure 8 Experimental ROC curves are shown to demonstrate the accuracy of the disclosed method for assessing IBD in subjects. Figure 8 As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.915, indicating that the disclosed method has an accuracy of over 90% for IBD determination based on tooth samples.

[0233] Table 8: Characteristics and β values ​​for IBD.

[0234]

[0235]

[0236]

[0237]

[0238]

[0239]

[0240]

[0241]

[0242]

[0243]

[0244]

[0245]

[0246]

[0247]

[0248] Example 7 – Kidney Transplant Rejection Prediction

[0249] Hair samples were collected from kidney transplant recipients with biopsy-confirmed acute rejection (n=6) and age- and sex-matched control kidney transplant recipients without acute rejection at biopsy during post-transplant monitoring (n=5). All participants were recruited from Mount Sinai Hospital. Table 9 shows the features used and their corresponding β values. β values ​​were obtained by estimating the change in log odds for each corresponding feature describing kidney transplant status in the corresponding cohort as a result of a one-unit change in the corresponding feature.

[0250] Figure 9 ROC curves are shown to illustrate the accuracy of the disclosed method for assessing kidney transplant rejection in subjects. Figure 9 As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.900, indicating that the disclosed method has 90% accuracy in assessing kidney transplant rejection based on hair samples.

[0251] Table 9: Characteristics and β values ​​for predicting kidney transplant rejection.

[0252]

[0253]

[0254]

[0255] Case 8 – Childhood Cancer

[0256] Pediatric cancer in subjects was assessed using methods similar to those described above for Examples 2 and 3. A total of 28 children were recruited from a hospital cancer center. Twenty-two were pediatric cancer cases, and six were controls. Diagnosis was made using standard clinical protocols, namely blood tests and histopathology, and the diagnosis was confirmed by an oncologist. Table 10 shows the features used and their corresponding beta values. Beta values ​​were obtained by estimating the change in log odds for each feature describing the pediatric cancer status in the corresponding cohort, which is associated with a one-unit change in the corresponding feature.

[0257] Figure 10 ROC curves are shown to illustrate the accuracy of the disclosed method for assessing pediatric cancer in subjects. Figure 10 As shown, the AUC of the ROC curve used to evaluate the performance of the disclosed classification method, derived from experimental data, corresponds to 0.962, indicating that the disclosed method has an accuracy of over 95% in pediatric cancers across a cohort of 28 children based on tooth sampling.

[0258] Table 10: Characteristics and β values ​​determined for pediatric cancers.

[0259]

[0260]

[0261]

[0262]

[0263]

[0264]

[0265]

[0266]

[0267]

[0268]

[0269] References cited and alternative embodiments

[0270] All references cited in this document are incorporated herein by full reference and for all purposes are as if each individual publication or patent or patent application were specifically and individually instructed to be incorporated herein by full reference for all purposes.

[0271] Many modifications and variations can be made to this invention without departing from its spirit and scope, as will be apparent to those skilled in the art. Specific embodiments described herein are given by way of example only. These embodiments were chosen and described to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to best utilize the invention and its various embodiments with modifications suitable for the contemplated particular purpose. This invention is defined only by the full scope of the appended claims together with their equivalents.

Claims

1. An apparatus for assessing biological symptoms related to metal metabolism in a subject, the apparatus comprising one or more processors and a memory storing one or more programs executed by the one or more processors, the one or more programs including instructions for performing the following: A laser beam is used to sample each of multiple locations along the longitudinal direction of the detached hair shaft sample, thereby obtaining multiple ion samples, each of the multiple ion samples corresponding to a different location among the multiple locations, and each of the multiple locations representing a different growth stage of the hair shaft sample. Each of the plurality of ion samples is analyzed using an inductively coupled mass spectrometer to obtain a first dataset containing multiple traces, each of the multiple traces being the concentration of a corresponding element isotope among multiple element isotopes determined together from the plurality of ion samples over time. A second dataset comprising a set of features is derived from the plurality of traces, each corresponding feature in the set of features being determined by variations in a single isotope or combination of isotopes in the plurality of traces; and The set of features is input into a trained classifier, thereby determining with at least 80% accuracy that the subject suffers from the biosymptom associated with metal metabolism, wherein the biosymptom associated with metal metabolism is autism spectrum disorder.

2. A non-transitory computer-readable storage medium storing instructions, when executed by a computer system, to cause the computer system to perform a method for assessing a subject's biological symptoms related to metal metabolism, the method comprising: A laser beam is used to sample each of a plurality of locations along the longitudinal direction of the detached hair shaft sample of the subject, thereby obtaining a plurality of ion samples, each of the plurality of ion samples corresponding to a different location among the plurality of locations, and each of the plurality of locations representing a different growth stage of the hair shaft sample. Each of the plurality of ion samples is analyzed using an inductively coupled mass spectrometer to obtain a first dataset containing multiple traces, each of the multiple traces being the concentration of a corresponding element isotope among multiple element isotopes determined together from the plurality of ion samples over time. A second dataset comprising a set of features is derived from the plurality of traces, each corresponding feature in the set of features being determined by variations in a single isotope or combination of isotopes in the plurality of traces; and The set of features is input into a trained classifier, thereby determining with at least 80% accuracy that the subject suffers from the biosymptom associated with metal metabolism, wherein the biosymptom associated with metal metabolism is autism spectrum disorder.

3. A classification method, comprising: In a computer system having one or more processors and memory storing one or more programs to be executed by said one or more processors: a) For each corresponding training subject among multiple training subjects, obtain the first dataset and the second dataset. The first subset of the training subjects among the plurality of training subjects has a first diagnostic state corresponding to having a first biological symptom related to metal metabolism, and the second subset of the training subjects among the plurality of training subjects has a second diagnostic state corresponding to not having the first biological symptom related to metal metabolism; The first dataset contains multiple corresponding traces. Each of these traces represents the concentration of a corresponding element isotope among multiple element isotopes, determined over time based on multiple corresponding ion samples. These multiple ion samples are obtained by sampling each of multiple corresponding positions along the longitudinal direction of a corresponding detached hair shaft sample from the corresponding training subject using a laser beam. Each ion sample corresponds to a different position among these multiple positions, and each position represents a different growth stage of the corresponding hair shaft sample. The multiple traces are obtained by analyzing each corresponding ion sample using an inductively coupled mass spectrometer. The second dataset contains a corresponding set of features, and each corresponding feature in the corresponding set of features is determined by the variation of a single isotope or combination of isotopes in the corresponding plurality of traces. as well as b) Using the following training of untrained or partially untrained classifiers, thereby obtaining a trained classifier, the trained classifier providing an indication with at least 80% accuracy on whether the test subject suffers from the first biological symptom related to metal metabolism based on the values ​​of features from a set of features obtained from a set of features obtained from the metal metabolism-related biological samples of the test subject: (i) the corresponding set of features for each of the corresponding second datasets of each of the plurality of training subjects; (ii) the corresponding diagnostic state of each of the plurality of training subjects, selected from the first diagnostic state and the second diagnostic state, wherein the first biological symptom related to metal metabolism is autism spectrum disorder.

4. The classification method according to claim 3, wherein the trained classifier is a neural network algorithm, a support vector machine algorithm, a decision tree algorithm, an unsupervised clustering model algorithm, a supervised clustering model algorithm, or a regression model.

5. The classification method according to claim 3, wherein the trained classifier is multinomial.

6. The classification method according to claim 3, wherein the trained classifier is binomial.

7. The classification method according to claim 3, wherein the plurality of element isotopes are selected from the following element isotopes: Li-7, Mg-24, Mg-25, Al-27, P-31, S-34, Ca-44, Ca-43, Cr-52, Mn-55, Fe-56, Co-59, Ni-60, Cu-63, Zn-66, As-75, Sr-88, Cd-111, Sn-118, I-127, Ba-138, Hg-201, Pb-208, Bi-209 and Mo-95.

8. The classification method according to claim 3, wherein each feature in the corresponding set of features is associated with a single corresponding trace or two corresponding traces in the corresponding plurality of traces.

9. The classification method according to claim 3, wherein the corresponding set of features comprises one or more features listed in the following group: entropy_Ni, determinism_Bi, stratification_Li, determinism_ZnSr, MDL_As, determinism_ZnAs, determinism_ZnCd, determinism_ZnS, determinism_ZnCa, determinism_ZnPb, determinism_ZnFe, determinism_S, entropy_Cu, entropy_Sr, entropy_Pb, entropy_Ca, MDL_Ni, MDL_Li, entropy_P, stratification_As, entropy_Cr, stratification_Mn, stratification_Cd, entropy_Co, stratification_Mg, entropy_Cd, entropy_Mg, TT_Pb, entropy_Sn, entropy_ZnCd, TT_P Layered Cu, TT_Zn, Layered Sn, MDL_P, MDL_ZnCd, Layered Fe, Layered Co, MDL_Pb, TT_As, MDL_Sr, MDL_Cd, MDL_Ca, Deterministic ZnLi, MDL_Cu, Layered Pb, Layered Bi, Entropy_Mn, MDL_Cr, MDL_Mg, TT_Mn, TT_S, MDL_Sn, Deterministic ZnAl, TT_Mg, MDL_ZnAs, RT2_Mn, TT_Li, TT_Sr, Entropy_ZnMn, MDL_Co, Deterministic Co, TT_Ca, TT_Cd, RT2_Ni, TT_Fe, RT2_Fe, MD L_ZnBi, RT2_ZnAl, RT2_Zn, RT2_Al, MDL_ZnMn, Layered_Zn, TT_Cu, MDL_ZnBa, RT2_P, RT2_ZnFe, MDL_Mn, RT2_Cr, Entropy_ZnBa, RT2_Cd, RT2_ZnS, RT2_S, RT2_Pb, RT2_ZnMn, MDL_ZnLi, RT2_ZnAs, Entropy_ZnAs, RT2_Sr, RT2_ZnSr, MDL_Zn, Layered_Ca, RT2_ZnCd, RT2_ZnLi, RT2_ZnSn, MDL_ZnMg, RT2_Sn, RT2_ZnMg, Entropy_ZnBi, R T2_ZnNi, MDL_ZnNi, RT2_ZnBi, RT2_Mg, RT2_Ba, RT2_ZnCu, RT2_ZnBa, RT2_ZnP, RT2_Co, RT2_ZnHg, RT2_Cu, RT2_ZnCo, stratification_Sr, RT2_Ca, RT2_ZnCa, MDL _ZnCa, RT2_ZnPb, entropy_Zn, RT2_Bi, MDL_I, entropy_I, stratification_Ni, MDL_ZnSr, MDL_ZnP, RT2_ZnCr, RT2_ZnI, MDL_ZnFe, RT2_As, entropy_ZnMg, MDL_ZnSn, TT_Al, MDL_ZnHg,Entropy_ZnSn, MDL_ZnCr, MDL_Ba, TT_Bi, RT2_Hg, Entropy_ZnP, MDL_ZnPb, TT_Sn, RT2_I, TT_Ba, TT_I, TT_Ni, MDL_ZnAl, MDL-Bil, RT2_Li, Entropy_ZnCo, Entropy_ZnLi, Entropy_ZnNi, Entropy_ZnCa, MDL_Fe, MDL_S, MDL_ZnCo, Entropy_ZnHg, TT_Co, MDL_ZnI, Entropy_ZnPb, MDL_Al, Entropy_ZnCr, Entropy_Ba, Entropy_ ZnFe, MDL_ZnS, MDL_Hg, Entropy_ZnSr, Entropy_S, TT_Hg, Stratification_Al, Entropy_ZnAl, Entropy_Bi, Determinism_ZnP, Entropy_Fe, Determinism_ZnBi, Entropy_ZnI, Stratification_Ba, Determinism_I, TT_Cr, Determinism_Ba, Stratification_I, Determinism_Sn, Determinism_ZnBa, Entropy_Al, Determinism_ZnCo, Entropy_Hg, Stratification_Cr, Stratification_P, Stratification_S, Determinism_Zn, Entropy_ZnS, Determinism_Al, Determinism_P and Stratification_Hg, in, The MDL is the average diagonal length, a key metric derived from recursive quantitative analysis. It reflects a simple measurement of the average length of the diagonals present in a two-dimensional recursive graph and serves as an absolute indicator of the duration of periodic components in a given signal. The recursive quantitative analysis is a nonlinear data analysis that quantifies the number and duration of recursions in a dynamic system. The certainty is the relative ratio of periodic components to aperiodic components in the recursive analysis; RT2 is the recursion time, which is the average time interval between diagonal elements, i.e., the interval between periods. The entropy is the variability of the distribution of the average diagonal length; TT is the capture time, which is the average length of the hierarchical structure in the two-dimensional recursive graph, and the hierarchical structure indicates a steady state. The hierarchical nature is an overall measure of signal stability.

10. The classification method according to claim 3, wherein the test subject is a human being.

11. The classification method according to claim 10, wherein the person is less than 5 years old.

12. The classification method according to claim 10, wherein the person is less than 1 year old.

13. The classification method according to claim 3, further comprising pretreating the corresponding hair shaft sample of the corresponding training subject with a solvent or surfactant prior to the sampling.

14. The classification method of claim 3, further comprising, prior to the sampling, irradiating the corresponding hair shaft sample of the corresponding training subject with a low-power laser to remove any debris from the corresponding hair shaft sample of the corresponding training subject.

15. The classification method of claim 3, wherein the sampling comprises irradiating the corresponding hair shaft sample of the corresponding training subject with the laser, thereby extracting a plurality of particles from the corresponding hair shaft sample of the corresponding training subject, and ionizing the plurality of particles with an inductively coupled plasma mass spectrometer, thereby obtaining the corresponding plurality of ion samples.

16. The classification method of claim 3, wherein the corresponding plurality of positions are arranged in order such that a first position among the plurality of positions along the corresponding hair shaft sample of the corresponding training subject corresponds to the position closest to the tip of the corresponding hair shaft sample of the corresponding training subject.

17. The classification method according to claim 3, wherein each of the corresponding plurality of traces comprises a plurality of data points, each data point being an instance of the corresponding position among the plurality of positions.

18. The classification method of claim 17, wherein deriving the second dataset comprises removing such data points from the plurality of data points that do not meet the first criterion.

19. The classification method of claim 18, wherein the first criterion comprises the mean absolute difference between adjacent data points in the corresponding plurality of data points being three times the standard deviation of the mean absolute difference between adjacent data points.

20. The classification method according to claim 3, wherein the concentration of the corresponding element isotope corresponds to the relative abundance of the corresponding element isotope relative to the control element isotope, the control element isotope being contained in the corresponding plurality of ion samples.

21. The classification method according to claim 20, wherein the reference element isotope is sulfur.

22. The classification method according to claim 3, wherein the corresponding set of features is selected from the group consisting of: average diagonal length, determinism, recursion time, entropy, capture time, and hierarchicality. in, The average diagonal length is a key metric derived from recursive quantitative analysis. It reflects a simple measurement of the average length of the diagonals present in a two-dimensional recursive graph and serves as an absolute indicator of the duration of periodic components in a given signal. The recursive quantitative analysis is a nonlinear data analysis that quantifies the number and duration of recursions in a dynamic system. The certainty is the relative ratio of periodic components to aperiodic components in the recursive analysis; The recursion time is the average time interval between diagonal elements, i.e., the interval between cycles; The entropy is the variability of the distribution of the average diagonal length; The capture time is the average length of the hierarchical structure in the two-dimensional recursive graph, which indicates a steady state. The hierarchical nature is an overall measure of signal stability.

23. The classification method according to claim 3, wherein the trained classifier calculates: in p(subject) is the probability that the test subject has the first biological symptom related to metal metabolism. e is the Euler number. α is when When equal to zero, it is a calculated parameter related to the probability that the test subject has the first biological symptom associated with metal metabolism. β 1,…,k The weight parameters correspond to the weights associated with each feature in the set of features from features 1 to k, and x 1,…,k The value corresponding to each feature in the set of features, which includes features 1 to k.

24. The classification method according to claim 3, wherein the corresponding plurality of positions comprises at least 100 positions.

25. The classification method according to claim 3, wherein the plurality of element isotopes comprises at least 22 element isotopes from the following: Li-7, Mg-24, Mg-25, Al-27, P-31, S-34, Ca-44, Ca-43, Cr-52, Mn-55, Fe-56, Co-59, Ni-60, Cu-63, Zn-66, As-75, Sr-88, Cd-111, Sn-118, I-127, Ba-138, Hg-201, Pb-208, Bi-209, and Mo-95.

26. A classification apparatus comprising one or more processors and a memory storing one or more programs executable by the one or more processors, the one or more programs including instructions for performing a classification method, the classification method comprising: a) For each corresponding training subject among a plurality of training subjects, wherein a first subset of the training subjects among the plurality of training subjects has a first diagnostic state corresponding to having a biological symptom related to metal metabolism, and a second subset of the training subjects among the plurality of training subjects has a second diagnostic state corresponding to not having said biological symptom related to metal metabolism: A laser beam is used to sample each of a plurality of corresponding locations along the longitudinal direction of the corresponding detached hair shaft sample of the corresponding training subject, thereby obtaining a plurality of corresponding ion samples, each of the plurality of corresponding ion samples corresponding to a different location among the plurality of corresponding locations, and each of the plurality of corresponding locations representing a different growth stage of the corresponding hair shaft sample. Each corresponding ion sample in the corresponding plurality of ion samples is analyzed by an inductively coupled mass spectrometer, thereby obtaining a corresponding first dataset containing multiple corresponding traces, each of the multiple corresponding traces being the concentration of a corresponding element isotope among multiple element isotopes determined together by the corresponding plurality of ion samples over time. A corresponding second dataset containing a corresponding set of features is derived from the corresponding plurality of traces, wherein each corresponding feature in the corresponding set of features is determined by the variation of a single isotope or combination of isotopes in the corresponding plurality of traces; as well as b) Using the following training of untrained or partially untrained classifiers, thereby obtaining a trained classifier, the trained classifier providing an indication of whether the test subject suffers from the metal metabolism-related biological condition with at least 80% accuracy based on the values ​​of features from a set of features obtained from metal metabolism-related biological samples of the test subject: (i) the corresponding set of features for each of the plurality of training subjects in each corresponding second dataset; (ii) the corresponding diagnostic state of each of the plurality of training subjects, selected from the first diagnostic state and the second diagnostic state, wherein the biological symptom associated with metal metabolism is autism spectrum disorder.

27. A non-transitory computer-readable storage medium storing instructions that, when executed by a computer system, cause the computer system to perform a classification method, the classification method comprising: a) For each corresponding training subject among a plurality of training subjects, wherein a first subset of the training subjects among the plurality of training subjects has a first diagnostic state corresponding to having a biological symptom related to metal metabolism, and a second subset of the training subjects among the plurality of training subjects has a second diagnostic state corresponding to not having said biological symptom related to metal metabolism: A laser beam is used to sample each of a plurality of corresponding locations along the longitudinal direction of the corresponding detached hair shaft sample of the corresponding training subject, thereby obtaining a plurality of corresponding ion samples, each of the plurality of corresponding ion samples corresponding to a different location among the plurality of corresponding locations, and each of the plurality of corresponding locations representing a different growth stage of the corresponding hair shaft sample. Each corresponding ion sample in the corresponding plurality of ion samples is analyzed by an inductively coupled mass spectrometer, thereby obtaining a corresponding first dataset containing multiple corresponding traces, each of the multiple corresponding traces being the concentration of a corresponding element isotope among multiple element isotopes determined together by the corresponding plurality of ion samples over time. A corresponding second dataset containing a corresponding set of features is derived from the corresponding plurality of traces, wherein each corresponding feature in the corresponding set of features is determined by the variation of a single isotope or combination of isotopes in the corresponding plurality of traces; as well as b) Using the following training of untrained or partially untrained classifiers, thereby obtaining a trained classifier, the trained classifier providing an indication of whether the test subject suffers from the metal metabolism-related biological condition with at least 80% accuracy based on the values ​​of features from a set of features obtained from metal metabolism-related biological samples of the test subject: (i) the corresponding set of features for each of the plurality of training subjects in each corresponding second dataset; (ii) the corresponding diagnostic state of each of the plurality of training subjects, selected from the first diagnostic state and the second diagnostic state, wherein the biological symptom associated with metal metabolism is autism spectrum disorder.

Images