Prediction of motor function scores in Pompe disease treatment

Abstract

A prediction method for forecasting changes in exercise function scores from baseline scores. The system acquires input data, including the time trajectory of a target indicator metabolite compound. It uses a prediction model and the input data to determine the predicted time trajectory of the exercise function score.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims priority to U.S. Provisional Patent Application No. 63 / 502,037, filed May 12, 2023, and U.S. Provisional Patent Application No. 63 / 624,935, filed January 25, 2024, both of whose disclosures are incorporated herein by reference in their entireties.

[0002] This specification generally relates to predicting the motor function of a subject based on biomarkers.

Background Art

[0003] This specification generally relates to predicting motor function scores for the treatment of Pompe disease.

[0004] Motor scores such as the Gross Motor Function Measure (GMFM) and the Quick Motor Function Test (QMFT) are used as valid endpoints for assessing motor function in the treatment of diseases that affect motor function. For example, GMFM and QMFT scores are used when evaluating the effectiveness of treatment for Pompe disease.

[0005] Pompe disease is a rare, autosomal - recessive disorder characterized by progressive, debilitating, and often fatal neuromuscular symptoms that affect multiple systems. It is caused by pathogenic variants of the GAA gene, resulting in a lack of acid α - glucosidase (GAA) enzyme activity and a progressive accumulation of glycogen in lysosomes.

[0006] Pompe disease is classified into two forms: infantile - onset Pompe disease (IOPD) and late - onset Pompe disease (LOPD). The disease presents as a spectrum from the perspective of onset and progression. In IOPD, symptoms typically appear before or at 12 months of age and include cardiomyopathy. In contrast, LOPD patients have symptom onset either after 12 months of age or before 12 months of age without cardiomyopathy.

[0007] Myozyme (alglucosidase alfa) is a globally approved enzyme replacement therapy (ERT) for the entire spectrum of Pompe disease, while Nexuviazyme (avalclucosidase alfa) is approved in the EU for LOPD and IOPD, but in the US it is approved only for LOPD.

[0008] Changes from baseline in GMFM-88% and / or QMFT total score have been used as efficacy endpoints to assess motor function in clinical trials in both IOPD and LOPD populations with Pompe disease. Changes from baseline in GMFM-88% and QMFT total score over time are influenced by the patient's initial functional status.

[0009] The urinary biomarker hexose tetrasaccharide (urinary Hex4 (mmol / mol)) may serve as an indirect measure of the degree of skeletal muscle glycogen clearance in Pompe disease. [Overview of the project] [Means for solving the problem]

[0010] This disclosure describes a method, computer system, and apparatus (including a computer program coded on a computer storage medium) for predicting motor function scores that measure one or more motor functions of a subject during treatment for Pompe disease.

[0011] In one embodiment, the disclosure provides a predictive method for predicting a change in an exercise function score from a baseline score. The exercise function score measures one or more exercise functions of a subject. The method may be implemented by a system including one or more computers. The system acquires input data including a time trajectory of an index metabolite compound of a subject. The time trajectory of the index metabolite compound includes the respective values ​​of the index metabolite compound of a subject at each of one or more time points. The system acquires a plurality of first model parameters of a predictive model and uses the input data to determine a predicted time trajectory of the exercise function score. Determining this includes, for each of one or more time points after the baseline time point, using the predictive model to determine the respective predicted values ​​of the change in the exercise function score from the baseline value at each time point, based on at least (i) the input data and (ii) the first model parameters. The predictive model includes a nonlinear function term of the change from the baseline value of the respective values ​​of the index metabolite compound at each time point. The system then outputs data characterizing the predicted time trajectory of the exercise function score.

[0012] In some implementations of the prediction method, the prediction model is a nonlinear random coefficient model containing one or more random parameters. In some cases, the rate of change parameter in the nonlinear function term is one of the random parameters. In some cases, to determine each prediction value, for each of the one or more random parameters, the system selects each value for each random parameter from its respective predefined distribution, and the prediction value is determined based on each selected value for the one or more random parameters.

[0013] In some implementations of the prediction method, the nonlinear function term is a decaying exponential function term. In some cases, the prediction model is a nonlinear random coefficient model, and the negative rate of change parameter in the exponential term is a random parameter. In some cases, the asymptote parameter of the exponential term is a fixed parameter.

[0014] In some implementations of the prediction method, the motor function score is the Gross Motor Function Scale (GMFM) score.

[0015] In some implementations of the prediction method, the motor function score is the upper limb function assessment test (QMFT) score.

[0016] In some implementations of the prediction method, the indicator metabolite compound is urinary hexose tetrasaccharide (uHex4), and the value of the indicator metabolite compound is the uHex4 concentration value. In some cases, one or more of the respective uHex4 concentration values ​​are obtained through measurements of the uHex4 concentration at each point in time for the subject.

[0017] In some implementations of the prediction method, the input data further includes (i) one or more demographic attributes of the subject, (ii) one or more clinical diagnoses of the subject, or (iii) the subject's treatment history.

[0018] In some implementations of the prediction method, each predicted value is determined based on (i) the baseline score of the subject, and (ii) the median of the baseline scores obtained for the subject population.

[0019] In some implementations of the prediction method, the predicted value is determined based on (i) the age of diagnosis of the subject, and (ii) the median age of diagnosis obtained for the subject population.

[0020] In some implementations of the prediction method, each predicted value is further determined based on the baseline median of an indicator metabolite compound obtained for the target population.

[0021] In some implementations of the prediction method, the subjects are those diagnosed with a specific disorder affecting motor function. In some cases, the specific disorder is infant-onset Pompe disease (IOPD). In other cases, the specific disorder is late-onset Pompe disease (LOPD).

[0022] In some implementations of the prediction method, the subjects are receiving enzyme replacement therapy for Pompe disease. In some cases, the therapy is alpha-glucosidase replacement therapy. In other cases, the therapy is avalglucosidase alpha replacement therapy.

[0023] In some implementations of the prediction method, obtaining the time trajectory of a target indicator metabolite compound involves obtaining a quantitative systems pharmacology (QSP) model of Pompe disease metabolism and, at each of one or more time points, using the QSP model to predict the time trajectory of the indicator metabolite compound's value as a biomarker of disease activity.

[0024] In some implementations of the prediction method, the subjects are simulation subjects that represent actual individuals diagnosed with a particular form of Pompe disease. In some cases, the form of Pompe disease is infancy-onset Pompe disease (IOPD). In other cases, the form of Pompe disease is late-onset Pompe disease (LOPD).

[0025] In some cases, the QSP model is used to simulate enzyme replacement therapy for the subject of simulation. In some cases, the enzyme replacement therapy is alpha-glucosidase replacement therapy. In some cases, the enzyme replacement therapy is avalglucosidase alpha replacement therapy.

[0026] In some implementations of the prediction method, the system further determines the therapeutic efficacy of enzyme replacement therapy based on at least the predicted time trajectory of the simulated motor function score. In some cases, the system further determines whether a particular real subject should receive enzyme replacement therapy based on at least the predicted time trajectory of each of the simulated motor function scores. In some cases, depending on the determination that a particular real subject should receive enzyme replacement therapy, the enzyme replacement therapy is physically administered to that real subject.

[0027] In some implementations of the prediction method, predicting the time trajectory of values ​​for an indicator metabolite compound using a QSP model involves obtaining a set of subject-specific model parameters for the QSP model and then using the QSP model to calculate the respective time trajectories of values ​​for the target indicator metabolite compound based on the set of subject-specific model parameters.

[0028] In some implementations of the prediction method, the target-specific set of model parameters is determined by adjusting one or more of the population-based set of model parameters based on one or more of the following: (i) one or more demographic attributes of the target, (ii) one or more clinical test results from the target, (iii) one or more clinical diagnoses of the target, or (iv) one or more treatment histories of the target. In some cases, the population-based set of model parameters is determined for the QSP model based on a first benchmark dataset of the target population. In some cases, the target population includes subjects diagnosed with IOPD. In some cases, the target population includes subjects diagnosed with LOPD.

[0029] In another aspect, the disclosure provides a parameter estimation method for determining model parameters of a predictive model for predicting changes in motor function scores from baseline scores. The parameter estimation method may be implemented by a system comprising one or more computers. The system obtains a second benchmark dataset for each of a plurality of subjects, comprising (i) respective input data specifying the respective values ​​of the change in an indicator metabolite compound of the subject at each of a plurality of time points, and (ii) respective output data specifying the respective output values ​​of the change in the motor function score of the subject at each of a plurality of time points. Based on the second benchmark dataset, the system determines model parameters of a predictive model, the predictive model including a nonlinear function term of the change in the value of an indicator metabolite compound of the subject, and outputs the model parameters.

[0030] In some implementations of parameter estimation methods, the motor function score is the Gross Motor Function Scale (GMFM) score.

[0031] In some implementations of parameter estimation methods, the motor function score is the upper limb function assessment test (QMFT) score.

[0032] In some implementations of the parameter estimation method, the indicator metabolite compound is urinary hexose tetrasaccharide (uHex4), and the value of the indicator metabolite compound is the uHEX4 concentration value.

[0033] In some implementations of parameter estimation methods, the predictive model is a nonlinear random coefficient model containing one or more random parameters. In some cases, the rate of change parameter in the nonlinear function term is one of the random parameters.

[0034] In some implementations of the parameter estimation method, the nonlinear function term is a decaying exponential function term. In some cases, the prediction model is a nonlinear random coefficient model, and the negative rate of change parameter in the exponential term is a random parameter. In some cases, the asymptote parameter of the exponential term is a fixed parameter.

[0035] In some implementations of the parameter estimation method, the predictive model further includes terms based on (i) the baseline score of the subject and (ii) the median of the baseline scores obtained for the subject population.

[0036] In some implementations of the parameter estimation method, the predictive model further includes terms based on (i) the age of diagnosis for the subject and (ii) the median age of diagnosis obtained for the subject population.

[0037] In some implementations of the parameter estimation method, the predictive model further includes a term based on the baseline median of an indicator metabolite compound obtained for the population under consideration.

[0038] The disclosure also provides a system comprising one or more computers and one or more storage devices that, when executed by one or more computers, store instructions causing one or more computers to perform the above-described method.

[0039] The disclosure also provides one or more computer storage media that, when executed by one or more computers, store instructions causing one or more computers to perform the above-described method.

[0040] The subject matter described herein may be implemented in specific embodiments to achieve one or more advantages.

[0041] The described technology provides a technique for predicting the motor function score of a subject over a period of time. For example, the described technology may enable the use of urinary hexose tetrasaccharide (uHex4), an indirect measure of the degree of skeletal muscle glycogen clearance in Pompe disease, as a biomarker to estimate improvement in a patient's motor function score under specific treatment.

[0042] The predictive technology offered can provide a cost-effective and time-efficient alternative to repetitive direct measurements of motor function, which may require extensive clinical assessments and / or specialized equipment. Measurement of uHex4 markers can be performed in a laboratory setting with a relatively fast turnaround time and lower cost compared to comprehensive motor function assessments. As a result, more frequent assessments can be performed during the treatment period, facilitating informed predictions of improvements in patients' motor function by researchers and clinicians, thus helping to guide treatment decisions and more effectively monitor disease progression.

[0043] In particular, several implementations of the described technologies combine quantitative systems pharmacology (QSP) models with predictive models to predict treatment outcomes for Pompe disease for different therapeutic approaches. QSP models can integrate diverse data, including data from clinical trials, real-world datasets, biological pathways, and individualized data from individual patients, along with mathematical representations of Pompe pathophysiology to simulate the enzyme deficiencies and resulting biomarker profiles that characterize Pompe disease. Predictive models link biomarkers (e.g., uHex4 concentration) to clinical outcomes (e.g., motor function indicators). The combination of QSP and predictive models provides a toolbox for personalized therapies and more effective drug development for Pompe disease.

[0044] For example, the technology provided enables a direct comparison of the results of different treatment options, such as aggregate glucosidase alfa and alglucosidase alfa, under controlled conditions for Pompe disease. Traditionally, comparing the effectiveness of different treatment options often involves clinical trials with inherent limitations such as patient heterogeneity and sample size constraints. The described technology uses a QSP model to create simulated patients that represent real individuals, enabling direct comparisons between different treatment options. This eliminates the need for additional physical testing, reduces patient risk, and potentially accelerates research efforts.

[0045] In another example, by creating "virtual twins" of actual patients, the described techniques can be used to predict how a particular patient might respond to each treatment option based on their unique demographics. This opens the door to personalized treatment plans, tailoring therapies to maximize individual benefits and minimize potential side effects. [Brief explanation of the drawing]

[0046] [Figure 1] This document shows the workflow of an example prediction system for predicting the motor function of a target individual. [Figure 2]An exemplary QSP model for describing cellular metabolism in Pompe disease is presented. [Figure 3] This flowchart shows an example of a process for predicting the motor function of a target. [Figure 4] This demonstrates the performance of the motor function score prediction model. [Figure 5] This demonstrates the performance of the QSP model. [Figure 6] This is a block diagram of an exemplary computer system. [Modes for carrying out the invention]

[0047] Similar reference numbers and designations in various drawings refer to the same elements.

[0048] Assessing motor function is crucial in several areas of medicine and research. For example, in diseases that cause muscle weakness or difficulty with movement, such as Pompe disease, evaluating the time trajectory of changes in motor function helps physicians assess the severity, progression, and / or effectiveness of treatment. Furthermore, assessing motor function is essential during the development of new treatments or interventions. In clinical trials, changes in motor function scores are often used as a primary outcome measure of treatment effectiveness. Accurate assessment and prediction help researchers design better trials, select appropriate sample sizes, and determine the optimal treatment intervention and timing for monitoring patients for significant outcomes.

[0049] While directly measuring motor function through clinical assessments is important, it can be inefficient and impractical for several reasons. Firstly, these assessments are time-consuming, may require specialized equipment and trained personnel, and limit how often they can be performed or how many patients can be monitored in clinical trials. Secondly, while direct measurements cannot be made for hypothetical treatments, researchers or clinicians may need to estimate how a patient might respond to a particular treatment or compare outcomes from multiple potential treatment options before applying the treatment.

[0050] This specification describes techniques for predicting motor function based on biomarkers such as urinary hexose tetrasaccharide (uHex4) in pump disease. The described predictive techniques provide a method for tracking disease progression and potentially estimating a patient's response to treatment without frequently using time-consuming and expensive clinical assessments of a patient's motor function.

[0051] Figure 1 shows the workflow of an example prediction system 100 for predicting the motor function of subject 110. The prediction system 100 is an example of a system implemented as a computer program on one or more computers in one or more locations where the systems, components, and technologies described below are implemented.

[0052] The prediction system 100 includes a prediction model 150. The prediction system 100 uses the prediction model 150 to process data including (i) target data 112 and (ii) the time trajectory of indicator metabolites 114 to predict the time trajectory of the target exercise function score 140. Specifically, at each time point after the baseline, the prediction model 150 predicts the change in the exercise function score from the baseline score based on the measurement of indicator metabolites at that time point.

[0053] The points of time used to track a time trajectory can be customized to suit specific research or clinical needs. The trajectory may cover days, weeks, months, or years, with points of time taken at intervals of days, weeks, or months accordingly.

[0054] A motor function score can be any appropriate metric that measures one or more motor functions of a subject. In some cases, the motor function score is a Gross Motor Function Scale (GMFM) score. For example, it could be a GMFM-88% total score or a GMFM-66% total score. In other cases, the motor function score can be a Qualitative Muscle Function Test (QMFT) score. Some other examples of motor function scores include the Time-Up-and-Go Test (TUG) score and the Nine-Hole Peg Test (NHPT) score.

[0055] The prediction system 100 outputs a predicted motor score trajectory 140. In some cases, the predicted motor score trajectory 140 may be used to monitor the progression of a disease affecting motor function. In some cases, the predicted motor score trajectory 140 can be used to determine the effectiveness of a treatment 170 administered to a subject 110 and to adjust the treatment strategy. For example, the prediction system 100 can be used to generate predicted motor score trajectories 140 for a subject 110 before and after the initiation of treatment. An increase in the predicted motor score after the subject receives treatment would suggest the effectiveness of the treatment.

[0056] In some cases, a predictive motor score trajectory 140 can be used to select a specific treatment 170 for subject 110. In certain cases, for patients diagnosed with Pompe disease, the predictive system 100 or another system can determine the therapeutic efficacy of enzyme replacement therapy (e.g., alpha-glucosidase replacement therapy or avalglucosidase alpha replacement therapy) based on at least the predictive motor score trajectory 140. In other cases, the predictive system 100 or another system can determine whether subject 110 should receive enzyme replacement therapy based on at least the predictive motor score trajectory 140. Depending on whether it is determined that subject 110 should receive enzyme replacement therapy 170, the treatment is administered physically to subject 110.

[0057] The subject-specific data 112 may include data that characterizes the subject 110, and may include data such as demographic attributes (e.g., age and sex), diagnostic history (e.g., age at diagnosis), treatment history (e.g., duration of previous treatment), and clinical data (e.g., body mass index, creatine kinase levels). The subject-specific data 112 may be part of the input to the exercise score prediction model 150.

[0058] The indicator metabolite trajectory 114 is the time trajectory of values ​​of indicator metabolite compounds that function as biomarkers of disease activity in diseases affecting motor function. For example, urinary biomarker hexose tetrasaccharide concentration (uHex4 (mmol / mol)) can function as an indirect measure of the degree of skeletal muscle glycogen clearance in Pompe disease, correlating with changes in motor function in subjects 110 with Pompe disease.

[0059] In some cases, the indicator metabolite trajectory 114 is obtained by laboratory measurement. For example, the uHex4 concentration value in the target urine sample can be measured for the subject 110 at multiple time points. The uHex4 concentration value can be measured using any appropriate measurement technique, such as tandem mass spectrometry (MS / MS) or enzyme assay. The exercise score prediction model 150 is configured to process the measured values ​​of the indicator metabolite compound at each time point to predict the subject's exercise function score at that time point.

[0060] Clinical assessment of motor function is time-consuming and requires specialized equipment and trained personnel. In contrast, laboratory measurement of indicator metabolite compounds is generally much more accessible. Therefore, using measured indicator metabolite compounds as biomarkers to predict a subject's motor function provides a more efficient and accessible means of monitoring changes in motor function in subject 110. As described above, the motor score trajectory 140 predicted by the measured indicator metabolite trajectory 114 can be used to monitor the progression of diseases affecting motor function, determine the effectiveness of treatment 170 administered to subject 110, and selectively adjust the treatment strategy.

[0061] In some cases, instead of obtaining an indicator metabolite trajectory 114 by laboratory measurements, system 100 or another system can predict the indicator metabolite trajectory 114 using a quantitative systems pharmacology (QSP) model 200 of disease metabolism for diseases affecting motor function. A specific example of a QSP model 200 for Pompe disease is illustrated with reference to Figure 2. Generally, a QSP model for Pompe disease simulates the key processes underlying the disease. This may represent how a deficiency in the GAA enzyme leads to glycogen accumulation in muscle cells and can predict changes in biomarkers such as uHex4 for a particular subject 110. The QSP model 200 can also simulate enzyme replacement therapy. Thus, system 100 or another system can use the QSP model to predict the uHex4 trajectory of a simulated subject representing an actual subject 110 diagnosed with a certain form of Pompe disease.

[0062] In particular, the QSP model 200 can be used to simulate enzyme replacement therapy (which is not actually administered to subject 110) and predict the subject's uHex4 trajectory under enzyme replacement therapy. Thus, the QSP model 200 can be used to simulate different treatment scenarios (e.g., different therapeutic agents, doses, and / or administration timelines) and predict the uHex4 trajectory for subject 110 under different treatment scenarios without subject 110 receiving any treatment. Subsequently, the exercise score trajectory 140 predicted from the simulated index metabolite trajectory (e.g., uHex4 trajectory) can be used to evaluate how the actual subject 110 responded to different treatments and to determine the choice of a particular treatment.

[0063] The exercise score prediction model 150 is configured to process the index metabolite values ​​at each time point after the baseline time point (e.g., the first time point) of the index metabolite trajectory 114 to generate respective predicted values ​​for the change in the exercise function score from the baseline value of the exercise function score. The exercise score prediction model 150 has a set of modal parameters 155.

[0064] Generally, the motor score prediction model 150 includes a nonlinear function term representing the change from the baseline value of each index metabolite compound at each time point. In the case of Pompe disease, the nonlinear function term correlates with the clinical observation of a nonlinear increase in motor function accompanied by a decrease in uHex4. For example, after the initiation of treatment, the change from baseline in motor function has been observed to continue improving over time, with greater improvement seen in earlier treatment periods. This change has been observed to reach a "plateau" after the stabilization of the uHex4 concentration change. In some cases, the nonlinear function term is a decaying exponential function term.

[0065] In some cases, the motion score prediction model 150 is a nonlinear random coefficient model that includes one or more random parameters. For example, the rate of change parameter in a nonlinear function term may be a random parameter. In these cases, the random parameters are defined by their respective distributions. To calculate the predicted motion score trajectory 140 for a specific object in a particular example, the random parameters are sampled from their respective predefined distributions, and the predicted motion score is calculated using the sampled values ​​of the random parameters.

[0066] In some cases, the nonlinear function term is a decaying exponential function term. The decay rate in the decaying exponential function term can be a random parameter, and the asymptote parameter of the exponential function term can be a fixed parameter.

[0067] The exact form of the exercise score prediction model 150 can be adjusted to describe a specific clinical scenario. In some cases, the predicted motor function score is further determined based on (i) the baseline score of the subject and (ii) the median of the baseline scores obtained for the population of subjects. In some cases, the predicted motor function score is further determined based on (i) the diagnostic age of the subject and (ii) the median of the diagnostic ages obtained for the population of subjects. In some cases, the predicted motor function score is further determined based on the baseline median of the indicator metabolite compounds obtained for the population of subjects. For example, subject 110 may be a subject diagnosed with a specific disease that affects motor function, and the population of subjects includes subjects diagnosed with a specific disease and under a specific treatment.

[0068] In an exemplary example, for a population (I) of subjects diagnosed with the IOPD form of Pompe disease, the exercise score prediction model 150 for predicting the change in the GMFM score (e.g., GMFM-88) of subject i at time point j may take the following form. △GMFM I,i,j =C I,o +CI ,1 (b.GMFM i -median I,gmfm )+C I,2 (b.uHex4 i -median I,uHex4 )+C I,3 (Age.diag i -median I,age )+C I,4 (b.Mdur i -median I.Mdur )+C I,5 (gender i )+(C I,6 +C I,7 *(b.Mdur i 0median I,Mdur ))×(1-e (-ratei*(-%△uHex4i,j)) )+ε ij

[0069] The parameter C I,0 ~C I,7 is a fixed parameter obtained for population I, and b.GMFMi This is the baseline value of the GMF score for subject i, and the median I,gmfm This is the median GMFM score for group I, and b.uhex4 i This is the baseline value of uHex4 for target i, and the median I,uHex4 This is the median of the uHex4 marker for population I, and Age.diag i This is the age of diagnosis for subject i, and the median I,age b.Mdur is the median age of diagnosis for group I. i This is the duration of previous myozyme (alglucosidase alpha) use for subject i at baseline, and rate i is, rate i A random parameter sampled from a first normal distribution > 0, where ε ij ΔuHex4 is another random parameter sampled from a second normal distribution. i,j This represents the change from the baseline value of uHex4 for subject i at time point J. The parameters (mean and standard deviation) defining the first and second normal distributions are determined based on the population I.

[0070] In another illustrative example, for a group of subjects (L) diagnosed with the LOPD form of Pompe disease, a motor score prediction model 150 for predicting the change in subject i's QMFT score at time j may take the following form: △QMFT I,i,j =C L,0 +C L,1 (b.QMFT i -median L,qmft )+C L,2 (b.uHex4 i -median L,uHex4 )+C L,3 (Age.diag i -median L,age )+C L,4 (age.infusion i -median L,ageinf )+C L,5 (b. BMI i -median L,bmi)+C L,6 (b.CK i -median L.CK )+C L,7 (b.CREAT i -median L,CREAT )+C L.8 ×(1-e(-rate i * (-%△uHex4i,j)) )+ε ij

[0071] Parameter C L,0 ~C L,8 These are fixed parameters obtained for the population L, and b.QMFT i This is the baseline value of the QMFT score for subject i, and the median I,QMFT This is the median QMFT score of group L, and b.uhex4 i This is the baseline value of uHex4 for target i, and the median L,uHex4 This is the median of the uHex4 markers in population L, and age.infusion i This is the age at which injections began for subject i, and the median L,ageinf b.BMI is the median age at initiation of infusion in population L. i This is the body mass index (BMI) of subject i at baseline, and the median L,bmi This is the median BMI of group L, and b.CK i This is the creatine kinase (CK) measurement value for subject i, and the median L,CK This is the median CK measurement value for population L, and b.CREAT i This is the creatine measurement value for subject i, and the median L,CREAT This is the median creatine measurement value for population L, and rate i is, rate i εij is a random parameter sampled from a third normal distribution where >0, and ΔuHex4 is another random parameter sampled from a fourth normal distribution. i,j This represents the change from the baseline value of uHex4 for object i at time j. The parameters defining the third and fourth normal distributions are determined based on the population L.

[0072] The parameters 155 of the exercise score prediction model 150 can be determined by a parameter estimation system 160 based on benchmark data 160 for a specific population of subjects, such as a population of patients diagnosed with a particular form of Pompe disease (e.g., IOPD or LOPD). The benchmark data 160 includes, for each set of subjects, (i) the respective input data of the model specifying the time trajectory of changes in the index metabolite compounds of each subject, and (ii) the respective output data specifying the time trajectory of changes in the exercise function score of each subject. The benchmark data may further include subject-specific data for each subject, such as demographic attributes, diagnostic history, treatment, and clinical data.

[0073] The parameter estimation system 160 can determine the model parameters of a predictive model based on a benchmark dataset using any appropriate technique. The parameter estimation system 160 can perform an optimization process to identify the optimal set of model parameters 155 that describe the relationship between biomarker changes and motor function changes in the benchmark dataset. In some cases, the motor score predictive model 150 is a nonlinear random coefficient model, and the model parameters 155 include parameters that describe fixed effects (applicable to the entire population) and parameters that describe the distribution of random effects (between individuals). In some cases, optimization may be performed using maximum likelihood estimation (MLE) to identify the model parameters that maximize the probability of observing the motor function changes observed in the benchmark data. MLE techniques often involve an iterative method that starts with an initial estimate and refines the parameters until a good fit is achieved. In some cases, optimization may be performed using a Bayesian method that starts with a prior distribution of model parameters and then updates the past by incorporating the benchmark data. The Bayesian approach can be computationally intensive but allows for greater flexibility and the ability to include prior knowledge. Examples of computational techniques include the steepest descent algorithm, the Newton-Raphson algorithm, and the expectation maximization (EM) algorithm. The choice of algorithm may depend on factors such as the complexity of the model, the size of the dataset, and the desired computation speed. In some cases, these techniques can be implemented using specialized statistical software.

[0074] Figure 2 shows an exemplary QSP model 200 for describing cellular metabolism in Pompe disease. In particular, the QSP model 200 represents molecular-level reactions in the cytoplasm 210 of a representative cell and is linked to the indicator metabolite hexose tetrasaccharide (Hex4) in plasma 240 and urine 230 via extracellular fluid 220.

[0075] The QSP200 model simulates a deficiency in acid alpha-glucosidase (GAA) activity that results in observed elevated glycogen storage and key urinary and plasma biomarkers (such as Hex4) in affected tissues. The QSP200 model further simulates the effects of enzyme replacement therapy (ERT) 250 (such as alglucosidase alfa and avalglucosidase alfa) that result in decreased tissue-specific glycogen loading and observation of Hex4 in urine and plasma.

[0076] The parameters of the QSP model 200 can be determined for a population diagnosed with a specific form of Pompe disease (IOPD or LOPD). Multiple data sources can be used to determine the QSP model parameters, including, for example, pharmacodynamic data, previous clinical trials of ERT for Pompe disease, and Pompe registries.

[0077] As illustrated with reference to Figure 1, the QSP model 200 can be used by a motor function prediction system to predict the uHex4 trajectory of a simulated subject representing a real subject diagnosed with a certain form of Pompe disease. In particular, the QSP model 200 can be used to simulate a specific enzyme replacement therapy for a simulated subject and predict the subject's uHex4 trajectory under enzyme replacement therapy. Thus, the QSP model 200 can be used to simulate different treatment scenarios (e.g., different therapeutic agents, doses, and / or timing of administration) and predict the uHex4 trajectory of a real subject under different treatment scenarios without the subject receiving treatment. The motor function prediction system can process the simulated uHex4 trajectories to predict the motor function score trajectory under these different treatment scenarios, which shows how the real subject responded to different treatments. This processing can be performed to assess the effect of different treatments on a population of subjects, taking into account heterogeneity across the patient cohort.

[0078] Figure 3 is a flowchart of an example process 300 for predicting the motor function of a subject. For convenience, process 300 is described as being performed by one or more computer systems located in one or more locations. For example, a training system, such as the motor function prediction system 100 of Figure 1, appropriately programmed according to this specification, can perform process 300.

[0079] In 310, the system acquires input data that includes the time trajectory of the target indicator metabolite compound. The time trajectory of the indicator metabolite compound includes the respective values ​​of the target indicator metabolite compound at each of one or more time points. Depending on the case, the input data may further include (i) one or more demographic attributes of the subject, (ii) one or more clinical diagnoses of the subject, or (iii) the subject's treatment history.

[0080] As described in more detail above with reference to Figure 1, the subjects may be those diagnosed with specific diseases affecting motor function, such as pump disease (IOPD or LOPD). The subjects may be receiving enzyme replacement therapy for Pompe disease, such as alpha-glucosidase replacement therapy or avalglucosidase alpha replacement therapy.

[0081] In some cases, the time trajectory of an indicator metabolite compound can be obtained using clinical tests. In some other cases, the time trajectory of an indicator metabolite compound can be generated using simulations, such as using a QSP model. In particular, as described in more detail above with reference to Figures 1 and 2, in some cases the subject is a simulation subject representing an actual subject diagnosed with IOPD or LOPD, and the uHex4 trajectory is predicted using a QSP model. In some cases, the QSP model is used to simulate enzyme replacement therapy (e.g., alpha-glucosidase replacement therapy or avalglucosidase alpha replacement therapy) for the simulation subject.

[0082] QSP model parameters for a specific subject can be determined by adjusting one or more of a set of population-based model parameters based on one or more of the following: (i) one or more demographic attributes of the subject, (ii) one or more clinical laboratory results from the subject, (iii) one or more clinical diagnoses of the subject, or (iv) one or more treatment history of the subject. As further explained above with reference to Figure 2, population-based model parameters can be obtained for a population diagnosed with a specific form of Pompe disease (IOPD or LOPD). Multiple data sources can be used to determine population-based QSP model parameters, including, for example, pharmacodynamic data, previous clinical trials of ERT for Pompe disease, and Pompe registries.

[0083] In 320, the system obtains several first model parameters for the motor function score prediction model. As described in more detail above with reference to Figure 1, the parameter estimation system can obtain first model parameters for the prediction model for each of several subjects based on a benchmark dataset that includes (i) respective input data specifying the time trajectory of the index metabolite compound for each subject, and (ii) respective output data specifying the time trajectory of the change in the motor function score for each subject.

[0084] In 330, the system determines the predicted time trajectory of the motor function score. Examples of motor function scores include the GMFM score and the QMFT score. Specifically, for each time point after the baseline in the time trajectory, the system uses a predictive model, based on at least (i) the input data and (ii) the first model parameters, to determine the respective predicted values ​​of the change in the motor function score from the baseline value at each time point.

[0085] Generally, a predictive model includes a nonlinear function term representing the change from the baseline value of each indicator metabolite compound at each point in time. As described in more detail above with reference to Figure 1, in some cases, the predictive model is a nonlinear random coefficient model that includes one or more random parameters. For example, the predictive model may include a decay exponential function term, where the negative rate of change parameter in the exponential term is a random parameter, while the asymptote parameter of the exponential term is a fixed parameter.

[0086] As described in more detail above with reference to Figure 1, in addition to the indicator metabolite trajectory, the predictive model further takes into account additional attributes of the subject population, such as (i) the subject's baseline score, (ii) the median baseline score obtained for the subject population, (iii) the subject's age at diagnosis, (iv) the median age at diagnosis obtained for the subject population, (v) the median baseline value of indicator metabolite compounds obtained for the subject population, (vi) the subject's treatment history, and (vii) additional clinical tests of the subject. The subject population includes subjects diagnosed with a specific disease.

[0087] In step 340, the system outputs data characterizing the predicted time trajectory of the motor function score. As described in more detail above with reference to Figure 1, the predicted motor function score trajectory can be used in various ways. For example, in some cases, the predicted motor function score trajectory can be used to determine the therapeutic effectiveness of enzyme replacement therapy for a subject. In some cases, the subject is a simulation subject corresponding to an actual subject, and the predicted motor function score trajectory of the simulation subject can be used to determine whether the actual subject should receive enzyme replacement therapy.

[0088] Figure 4 shows the performance of the motor function score prediction model. The motor function score prediction model was used to predict the GMFM-88 trajectory of three cohorts of patients under different ERT therapies prior to week 25. All patients were switched to avalglucosidase alfa 40 mg / kg after week 25. The predicted GMFM-88 trajectory was compared to the observed GMFM-88 trajectory. In general, the predicted GMFM-88 trajectory was consistent with the trend of the observed trajectory, demonstrating the performance of the motor function score prediction model.

[0089] Figure 5 shows the performance of the QSP model. For a group of patients, the QSP model is used to compare predicted uHex4 concentrations with observed uHex4 measurements. The comparison shows that the predicted uHex4 generally agrees with the observed uHex4, demonstrating the performance of the QSP model.

[0090] Figure 6 is a block diagram of an example computer system 600 that can be used to perform the operations described above. System 600 includes a processor 610, memory 620, storage device 630, and input / output device 640. Each of the components 610, 620, 630, and 640 can be interconnected, for example, using a system bus 650. The processor 610 can process instructions to be executed within system 600. In one implementation, the processor 610 is a single-threaded processor. In another implementation, the processor 610 is a multi-threaded processor. The processor 610 can process instructions stored in memory 620 or storage device 630.

[0091] Memory 620 stores information within the system 600. In one implementation, memory 620 is a computer-readable medium. In another implementation, memory 620 is a volatile memory unit. In yet another implementation, memory 620 is a non-volatile memory unit.

[0092] The storage device 630 can provide the system 600 with large-capacity storage. In one implementation, the storage device 630 is a computer-readable medium. In various different implementations, the storage device 630 may include, for example, a hard disk device, an optical disk device, a storage device shared over a network by multiple computing devices (e.g., cloud storage), or several other mass storage devices.

[0093] The input / output device 640 provides input / output operations for the system 600. In one implementation, the input / output device 640 may include one or more network interface devices, such as an Ethernet card, a serial communication device, such as an RS-232 port, and / or a wireless interface device, such as a 502.11 card. In another implementation, the input / output device may include a driver device configured to receive data and send output data to other input / output devices, such as a keyboard, printer, and display device 660. However, other implementations such as mobile computing devices, mobile communication devices, and set-top box television client devices may also be used.

[0094] An exemplary processing system is shown in Figure 6, but the subject matter and functional operations described herein may be implemented in other types of digital electronic circuits or computer software, firmware, or hardware, or one or more combinations thereof, including the structures disclosed herein and their structural equivalents.

[0095] This specification uses the term “configured” in relation to system and computer program components. In the case of a system of one or more computers, being configured to perform a particular operation or action means that software, firmware, hardware, or a combination thereof, that causes the system to perform an operation or process is installed on the system. One or more computer programs being configured to perform a particular operation or action means that, when executed by a data processing device, the programs include instructions that cause the device to perform an operation or process.

[0096] The subject matter and functional operating embodiments described herein may be implemented in digital electronic circuits, tangibly implemented computer software or firmware, computer hardware including the structures disclosed herein and their structural equivalents, or one or more combinations thereof. Embodiments of the subject matter described herein may be implemented as one or more modules of computer program instructions, i.e., computer program instructions executed by a data processing device or coded in a tangible non-temporary storage medium for controlling the operation of a data processing device. The computer storage medium may be a machine-readable storage device, a machine-readable storage board, a random or serial access memory device, or one or more combinations thereof. Alternatively or additionally, the program instructions may be coded in artificially generated propagating signals, such as mechanically generated electrical, optical or electromagnetic signals generated to code information for transmission to a receiving device suitable for execution by a data processing device.

[0097] The term "data processing device" refers to data processing hardware and encompasses all kinds of devices, equipment, and machines for data processing, including, for example, programmable processors, computers, or multiple processors or computers. A device may also be, or further include, dedicated logic circuits, such as FPGAs (Field-Programmable Gate Arrays) or ASICs (Application-Specific Integrated Circuits). Optionally, in addition to hardware, a device may include code that forms the execution environment for computer programs, such as processor firmware, protocol stacks, database management systems, operating systems, or one or more combinations thereof.

[0098] Computer programs, also called or written as programs, software, software applications, apps, modules, software modules, scripts, or code, may be written in any form of programming language, including compiler or interpreter languages ​​or declarative or procedural languages, and may be deployed in any form, including standalone programs or modules, components, subroutines, or other units suitable for use in a computer environment. A program may, but may not, correspond to a file in a file system. A program may be stored in part of a file that holds other programs or data, such as one or more scripts stored in a markup language document, a single file or multiple collaborative files dedicated to the program in question, such as a file that stores one or more modules, subprograms, or parts of code. A computer program may be deployed to run on one computer or one location, or distributed across multiple locations and interconnected by a data communication network.

[0099] In this specification, the term “engine” is used broadly to refer to a software-based system, subsystem, or process programmed to perform one or more specific functions. Generally, an engine is implemented as one or more software modules or components and installed on one or more computers located in one or more locations. In some cases, one or more computers are dedicated to a particular engine, while in other cases, multiple engines can be installed and operated on one or more of the same computers.

[0100] The processes and logic flows described herein may be executed by one or more programmable computers running one or more computer programs to perform functions by acting on input data and producing outputs. The processes and logic flows may also be executed by dedicated logic circuits, such as FPGAs or ASICs, or by a combination of dedicated logic circuits and one or more programmed computers.

[0101] A computer suitable for running computer programs may be based on a general-purpose or dedicated microprocessor, or both, or any other type of central processing unit. Generally, the central processing unit receives instructions and data from read-only memory, random-access memory, or both. Essential elements of a computer are a central processing unit that executes or carries out instructions, and one or more memory devices that store instructions and data. The central processing unit and memory may be reinforced or incorporated by dedicated logic circuits. Generally, a computer will be configured to store data in one or more mass storage devices, including, for example, magnetic, magneto-optical disks, or optical disks, or to receive data from or transfer data to them, or both. However, a computer is not required to have such devices. Furthermore, a computer may be incorporated into another device, to name a few, such as a mobile phone, personal digital assistant (PDA), mobile voice or video player, game console, Global Positioning System (GPS) receiver, or portable storage device, such as a Universal Serial Bus (USB) flash drive.

[0102] Computer-readable media suitable for storing computer program instructions and data include, for example, semiconductor memory devices such as EPROM and EEPROM, flash memory devices, magnetic disks such as internal hard disks or removable disks, magneto-optical disks, and all forms of non-volatile memory, media, and memory devices, including CD-ROM and DVD-ROM disks.

[0103] To enable user interaction, embodiments of the subject matter described herein may be implemented in a computer having a display device that displays information to the user, such as a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, and an indicator device that allows the user to provide input to the computer, such as a keyboard and mouse or trackball. Other types of equipment may be used to enable user interaction; for example, the feedback provided to the user may be any form of sensory feedback, such as visual feedback, auditory feedback, or tactile feedback, and input from the user may be received in any form, including acoustic, voice, or tactile input. In addition, the computer can interact with the user by sending and receiving documents to and from the equipment used by the user, for example, by sending a web page to a web browser on the user's equipment in response to a request received from a web browser. The computer can also interact with the user by sending text messages or other forms of messages to a personal device, such as a smartphone running a messaging application, and then receiving response messages from the user.

[0104] Data processing devices that implement machine learning models may include, for example, dedicated hardware accelerator units that handle the general and computationally intensive parts of machine learning training or production, i.e., inference workloads.

[0105] Machine learning models can be implemented and deployed using machine learning frameworks, such as the TensorFlow framework or the Jax framework.

[0106] Multiple embodiments of the subject matter described herein may be implemented in a computing system that includes, for example, a backend component as a data server, or a middleware component, such as an application server, or a frontend component, such as a client computer equipped with a graphical user interface, a web browser, or an application that allows a user to interact with the implementation of the subject matter described herein, or any combination of one or more such backend, middleware, or frontend components. The components of the system may be interconnected by digital data communication in any form or medium, such as a communication network. Examples of communication networks include local area networks (LANs) and wide area networks (WANs), such as the Internet.

[0107] A computing system may include a client and a server. The client and server are generally located remotely from each other and typically interact via a communication network. The relationship between the client and the server arises from computer programs running on each computer that have a client-server relationship with each other. In some embodiments, the server transmits data, such as an HTML page, to a user device for the purpose of displaying data to a user interacting with a device acting as a client and receiving input from the user. Data generated on the user device, such as the results of the interaction with the user, may be received from the device on the server side.

[0108] This specification includes many specific implementation details, but these should not be construed as limiting the scope of any invention or claim, but rather as descriptions of features specific to a particular embodiment of a particular invention. Certain features described herein in relation to separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in relation to a single embodiment may also be implemented in multiple embodiments separately or in any suitable partial combination. Furthermore, features may be described as operating in a particular way in combination, and may be claimed as such from the outset, but one or more features from the claimed combination may, in some cases, be excluded from that combination, and the claimed combination may refer to a partial combination or a variation of a partial combination.

[0109] Similarly, while operations are shown in the drawings and enumerated in the claims in a specific order, it should not be understood that such operations must be performed in a specific illustrated or sequential order, or that all illustrated operations must be performed, in order to achieve the desired result. In certain situations, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged in multiple software products.

[0110] This document describes specific embodiments of the subject matter. Other embodiments are also included in the scope of the following claims. For example, the desired results may be achieved by performing the operations mentioned in the claims in a different order. As an example, the processes shown in the accompanying drawings do not necessarily have to be performed in the specific illustrated or sequential order in order to obtain the desired results. Multitasking and parallel processing may be advantageous in some cases.

Claims

1. A computer implementation method for predicting motor function scores that measure one or more motor functions of a subject, The method involves obtaining input data including the time trajectory of the target indicator metabolite compound, wherein the time trajectory of the indicator metabolite compound includes, for each of one or more time points, the respective values ​​of the target indicator metabolite compound at each of those time points. Obtaining multiple first model parameters of the prediction model, Determining the predicted time trajectory of the motor function score using the input data, comprising, for each of the one or more time points after the baseline time point, determining, using the prediction model, at least (i) the input data and (ii) the first model parameters, each predicted value of the change in the motor function score from the baseline value at each of the time points, wherein the prediction model includes a nonlinear function term of the change from the baseline value of each of the values ​​of the index metabolite compounds at each of the time points. Outputting data that characterizes the predicted time trajectory of the motor function score. Computer implementation methods including

2. The method according to claim 1, wherein the prediction model is a nonlinear random coefficient model including one or more random parameters.

3. The method according to claim 2, wherein the rate of change parameter in the nonlinear function term is one of the random parameters.

4. Determining each of the aforementioned predicted values ​​includes, for each of the one or more aforementioned random parameters, selecting the respective values ​​for each of the aforementioned random parameters from the respective predefined distributions. The method according to claim 2, wherein the predicted value is determined based on the respective values ​​selected for one or more random parameters.

5. The method according to any one of claims 1 to 4, wherein the nonlinear function term is a decaying exponential function term.

6. The method according to claim 5, wherein the prediction model is a nonlinear random coefficient model, and the negative rate of change parameter in the exponential term is a random parameter.

7. The method according to claim 6, wherein the asymptote parameter of the exponential function term is a fixed parameter.

8. The method according to any one of claims 1 to 7, wherein the motor function score is a gross motor function scale (GMFM) score.

9. The method according to any one of claims 1 to 8, wherein the motor function score is the upper limb function evaluation test (QMFT) score.

10. The method according to any one of claims 1 to 9, wherein the indicator metabolite compound is urinary hexose tetrasaccharide (uHex4), and the value of the indicator metabolite compound is the uHex4 concentration value.

11. The method according to claim 10, wherein one or more of the aforementioned uHex4 concentration values ​​are obtained by measuring the uHex4 concentration of the subject at each of the aforementioned time points.

12. The method according to any one of claims 1 to 11, wherein the input data further comprises (i) one or more demographic attributes of the subject, (ii) one or more clinical diagnoses of the subject, or (iii) the treatment history of the subject.

13. The method according to any one of claims 1 to 12, wherein each of the aforementioned predicted values ​​is determined based on (i) the baseline score of the subject and (ii) the median of the baseline scores obtained for the subject population.

14. The method according to any one of claims 1 to 13, wherein the predicted value is determined further based on (i) the age of diagnosis of the subject and (ii) the median age of diagnosis obtained for the subject group.

15. The method according to any one of claims 1 to 14, wherein each of the aforementioned predicted values ​​is further determined based on the baseline median of the indicator metabolite compound obtained for the target population.

16. The method according to any one of claims 1 to 15, wherein the subject is a subject diagnosed with a specific disease affecting motor function.

17. The method according to claim 16, wherein the specific disease is infant-onset Pompe disease (IOPD).

18. The method according to claim 16, wherein the specific disease is late-onset Pompe disease (LOPD).

19. The method according to claim 16, wherein the subject is receiving enzyme replacement therapy for Pompe disease.

20. The method according to claim 19, wherein the therapy is alpha-glucosidase replacement therapy.

21. The method according to claim 19, wherein the therapy is avalglucosidase alpha replacement therapy.

22. Obtaining the time trajectory of the aforementioned target indicator metabolite compound is, To obtain a quantitative systems pharmacology (QSP) model of Pompe disease metabolism, At each of the one or more time points mentioned above, the QSP model is used to predict the time trajectory of the value of the indicator metabolite compound as a biomarker of disease activity. A method according to any one of claims 1 to 21, including the method described in any one of claims 1 to 21.

23. The method according to claim 22, wherein the subject is a simulation subject representing an actual subject diagnosed with a certain form of Pompe disease.

24. The method according to claim 23, wherein the form of Pompe disease is infancy-onset Pompe disease (IOPD).

25. The method according to claim 23, wherein the form of Pompe disease is late-onset Pompe disease (LOPD).

26. The method according to any one of claims 23 to 25, wherein the QSP model is used to simulate enzyme replacement therapy for the subject of the simulation.

27. The method according to claim 26, wherein the enzyme replacement therapy is alpha-glucosidase replacement therapy.

28. The method according to claim 26, wherein the enzyme replacement therapy is aggregate glucosidase alpha replacement therapy.

29. The method according to any one of claims 26 to 28, further comprising determining the therapeutic effectiveness of the enzyme replacement therapy based on the predicted time trajectory of at least the motor function score being simulated.

30. The method according to any one of claims 26 to 29, further comprising determining whether a particular actual subject should receive the enzyme replacement therapy based on at least the predicted time trajectories of the motor function scores of the simulated subject.

31. The method according to claim 30, further comprising physically administering the enzyme replacement therapy to the specific actual subject in response to the determination that the specific actual subject should receive the enzyme replacement therapy.

32. Using the QSP model to predict the time trajectory of the values ​​of the indicator metabolite compound is, Obtaining a set of model parameters specific to the target for the aforementioned QSP model, Based on the set of model parameters specific to the subject, the QSP model is used to calculate the time trajectory of the respective values ​​of the indicator metabolite compounds of the subject. The method according to any one of claims 22 to 31, including the method described in any one of claims 22 to 31.

33. The method according to claim 32, wherein the set of model parameters specific to the subject is determined by adjusting one or more of the population-based set of model parameters based on one or more of the following: (i) one or more demographic attributes of the subject, (ii) one or more clinical test results from the subject, (iii) one or more clinical diagnoses of the subject, or (iv) one or more treatment histories of the subject.

34. The method according to claim 33, wherein the set of population-based model parameters is determined for the QSP model based on a first benchmark dataset of the population of interest.

35. The method according to claim 34, wherein the target group includes subjects diagnosed with IOPD.

36. The method according to claim 34, wherein the target group includes subjects diagnosed with LOPD.

37. A computer implementation method for determining model parameters of a predictive model for predicting changes in motor function scores from baseline scores, wherein the motor function score measures one or more motor functions of a subject, and the method is For each of the multiple subjects, a second benchmark dataset is obtained, which includes (i) input data specifying the respective values ​​of the change in the subject's indicator metabolite compound at each of the multiple time points, and (ii) output data specifying the respective output values ​​of the change in the subject's exercise function score at each of the multiple time points. The determination of the model parameters of the prediction model based on the second benchmark dataset, wherein the prediction model includes a nonlinear function term of the change in the value of the target indicator metabolite compound. Outputting the aforementioned model parameters Computer implementation methods, including those mentioned above.

38. The method according to claim 37, wherein the motor function score is a gross motor function scale (GMFM) score.

39. The method according to claim 37, wherein the motor function score is the upper limb function evaluation test (QMFT) score.

40. The method according to any one of claims 37 to 39, wherein the indicator metabolite compound is urinary hexose tetrasaccharide (uHex4), and the value of the indicator metabolite compound is the uHex4 concentration value.

41. The method according to any one of claims 37 to 40, wherein the prediction model is a nonlinear random coefficient model including one or more random parameters.

42. The method according to claim 41, wherein the rate of change parameter in the nonlinear function term is one of the random parameters.

43. The method according to any one of claims 37 to 42, wherein the nonlinear function term is a decaying exponential function term.

44. The method according to claim 43, wherein the prediction model is a nonlinear random coefficient model, and the negative rate of change parameter in the exponential term is a random parameter.

45. The method according to claim 44, wherein the asymptote parameter of the exponential function term is a fixed parameter.

46. The method according to any one of claims 37 to 45, wherein the predictive model further includes a term based on (i) the baseline score of the subject and (ii) the median of the baseline scores obtained for the subject population.

47. The method according to any one of claims 37 to 46, wherein the predictive model further includes terms based on (i) the age of diagnosis of the subject and (ii) the median age of diagnosis obtained for the subject group.

48. The method according to any one of claims 37 to 47, further comprising a term based on the baseline median of the indicator metabolite compound obtained for the target population.

49. One or more computers, When executed by the one or more computers, the one or more computers store one or more storage devices that store instructions causing the one or more computers to perform the operation of each of the methods described in any one of claims 1 to 48. A system that includes this.

50. One or more computer-readable storage media that, when executed by one or more computers, store instructions causing one or more computers to perform the operation of each of the methods described in any one of claims 1 to 48.

Images