Method for determining the life of at least one chromatography column - Patents.com

Abstract

A computer-implemented method (140) for determining a lifetime of at least one chromatography column (116) of at least one chromatography device (110), comprising: i) receiving model input chromatography data via at least one communication interface (128), ii) using at least one processing device (130) to determine at least one state variable representative of the lifetime of the chromatography column (116) using at least one data-driven model based on the model input chromatography data, and iii) using the processing device (130) to determine information regarding the lifetime by evaluating the determined state variable, the evaluation comprising comparing the determined state variable to at least one threshold. Further disclosed are a test system (112), a computer program, and a method for operating the chromatography column (116).

Description

[Technical field]

[0001] The present invention relates to a method for determining the lifetime of at least one chromatography column of at least one chromatography apparatus, a test system configured to carry out this method, a method for operating a chromatography column, and a computer program. [Background technology]

[0002] In Vitro Diagnostic (IVD) instruments may use chromatographic separation, such as high performance liquid chromatography (HPLC) or rapid liquid chromatography (Rapid LC), of analytes before measuring them by mass spectrometry (MS). Such Rapid LC or HPLC columns are usually replaced periodically by the customer. The expected life span may depend on the respective use case. For example, the current expected life span is roughly 5000 injections, taking into account column degradation and contamination. However, due to the high cost of columns, they account for a large part of the overall assay cost. To ensure uninterrupted workflow and throughput of samples in the laboratory, columns need to be replaced before they fail. The gradual deterioration of HPLC or Rapid LC columns in liquid chromatography (LC) / MS systems is a naturally occurring phenomenon. Column failure can lead to poor measurement performance or even result in erroneous measurements, which is especially critical in the in vitro diagnostic (IVD) environment.

[0003] Since column life is influenced by random factors, unexpected column replacement events may occur, which is undesirable because the system downtime reduces throughput. To reduce the occurrence of such unexpected events, more frequent replacement may be performed, but premature replacement is also undesirable because columns are expensive items. To address both unexpected column failures and the lowest possible frequency of replacement, prediction of column life would be beneficial. However, traditional statistical methods or simple counters are expected to work poorly because they ignore individual columns, chromatographic equipment and laboratory, or environmental factors.

[0004] US Patent No. 5,670,379 describes a chromatographic system that establishes a regression line between the retention times of a given peak measured in each previous run for a standard sample having known components, with the peak identification conditions, i.e., time windows, being corrected with reference to this regression line.

[0005] Issues to be resolved It would therefore be desirable to provide a method and system for determining the life of at least one chromatography column of at least one chromatography device that at least partially addresses the above-mentioned technical problems, in particular, a method and system for determining the life of at least one chromatography column of at least one chromatography device that allows for increased reliability and accuracy, thus optimizing throughput and patient safety, while also allowing for reduced costs. Summary of the Invention

[0006] overview This problem is addressed by a method and a test device having the features of the independent claims. Advantageous embodiments, which may be implemented alone or in any combination, are set out in the dependent claims as well as in the entire specification.

[0007] When used below, the terms "having", "comprises" or "includes" or any grammatical variants thereof are used in a non-exclusive manner. These terms may therefore refer both to the situation where, apart from the features introduced by these terms, no further features are present in the entity described in this context, and to the situation where one or more further features are present. As an example, the expressions "A has B", "A comprises B" and "A includes B" may all refer to the situation where, apart from B, no other elements are present in A (i.e., A is exclusively composed of B), and to the situation where, apart from B, one or more further elements are present in entity A, such as element C, elements C and D, or even further elements.

[0008] Furthermore, it should be noted that the terms "at least one" or "one or more" or similar expressions indicating that a feature or element may be present one or more times are typically used only once when introducing each feature or element. In the following, in most cases, when referring to each feature or element, the expressions "at least one" or "one or more" will not be repeated, despite the fact that each feature or element may be present one or more times.

[0009] Furthermore, when used hereinafter, the terms "preferably", "more preferably", "in particular", "even more particularly", "specifically", "more specifically", or similar terms are used with respect to any feature without limiting the possibility of alternatives. Thus, the features introduced by these terms are any features and are not intended to limit the technical scope of the claims in any way. The present invention may be practiced, as would be understood by one of ordinary skill in the art, by using alternative features. Similarly, features introduced by "in an embodiment of the present invention" or similar expressions are intended to be any features and are not accompanied by any limitation regarding alternative embodiments of the present invention, nor any limitation regarding the technical scope of the present invention, and are not accompanied by any limitation regarding the possibility of combining such introduced features with any other or non-arbitrary features of the present invention.

[0010] As used herein, the term "standard conditions" refers, unless otherwise specified, to IUPAC standard ambient temperature and pressure (SATP) conditions, i.e., preferably, a temperature of 25° C. and an absolute pressure of 100 kPa, and also preferably, the standard conditions include a pH of 7. Furthermore, unless otherwise indicated, the term "about" refers to the indicated value with the technical precision generally accepted in the relevant field, preferably to the indicated value ±20%, more preferably ±10%, and most preferably ±5%. Furthermore, the term "essentially" means that there is no deviation that affects the indicated result or use, i.e., if deviations exist, the indicated result does not deviate more than ±20%, more preferably ±10%, and most preferably ±5%. Thus, "consisting essentially of" means that it includes the specified components, but excludes other components, except for materials present as impurities, unavoidable materials present as a result of the process used to bring about those components, and components added for purposes other than achieving the technical effect of the invention. For example, a composition defined using the phrase "consisting essentially of" includes any known acceptable additives, excipients, diluents, carriers, etc. Preferably, a composition consisting essentially of a set of components will contain less than 5% by weight, more preferably less than 3% by weight, even more preferably less than 1% by weight, and most preferably less than 0.1% by weight of unspecified components.

[0011] The method described herein is an in vitro method. The method, such as at least one step or all method steps, may be assisted or performed by automated equipment. In particular, the entire method may be performed in such an automated equipment, for example, in a chromatographic analysis system. The steps described may be performed in any order, as far as technically possible, but in further embodiments, are performed in a given order. Furthermore, the method may include steps other than those explicitly described above.

[0012] In a first aspect of the present invention, a computer-implemented method for determining the life of at least one chromatography column of at least one chromatography apparatus is proposed.

[0013] As used herein, the term "computer-implemented method" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically, but not limited to, refer to a method involving at least one computer and / or at least one computer network. The computer and / or computer network may comprise at least one processor configured to perform at least one of the method steps of the method according to the invention. Preferably, each of the method steps is performed by the computer and / or computer network. The method may be performed fully automatically, in particular without requiring user interaction. As used herein, the term "automatically" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically, but not limited to, refer to a process performed fully by at least one computer and / or computer network and / or machine, in particular without requiring manual action and / or user interaction.

[0014] As used herein, the term "chromatography column" is a broad term and should be given its usual and customary meaning to those skilled in the art, and should not be limited to a specific or special meaning. The term may specifically refer to, but is not limited to, a typically cylindrical container that contains a stationary phase and has an inlet and an outlet for a mobile phase, such as a liquid, gas, aqueous chromatography solvent, etc. For example, a chromatography column is a liquid chromatography (LC) column. For example, a chromatography column is a high performance liquid chromatography (HPLC) or fast performance chromatography (FPLC) or high performance liquid chromatography column. Suitable stationary phase materials and mobile phases and their combinations are known in the art.

[0015] The chromatography device may, for example, comprise at least one liquid chromatography device. The liquid chromatography device may be or comprise at least one high performance liquid chromatography (HPLC) device or at least one micro liquid chromatography (μLC) device. The chromatography device may be coupled to a mass spectrometry device, for example, via at least one interface. As used herein, the term "chromatography device" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a specific or special meaning. The term may specifically refer to, but is not limited to, an analytical module configured to separate one or more analytes of interest of a sample from other components of the sample for detection of the one or more analytes by a mass spectrometry device. The chromatography device may comprise at least one chromatography column. For example, the chromatography device may be a single column device or a multi-column device having multiple columns. The chromatography column may have a stationary phase through which a mobile phase is pumped to separate and / or elute and / or transfer analytes of interest. As used herein, the term "mass analyzer" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a specific or special meaning. The term may specifically refer to, but is not limited to, a mass analyzer configured to detect at least one analyte based on mass-to-charge ratio. The mass analyzer may be or include at least one quadrupole mass analyzer. The interface coupling the chromatography device and the mass analyzer may include at least one ionization source configured to generate molecular ions and transfer the molecular ions to the gas phase.

[0016] As used herein, the term "analyte" refers to any compound or group of compounds to be determined in a sample. For example, the analyte may be a macromolecule, i.e. a compound with a molecular weight of more than 1000u (i.e. more than 1kDa). For example, the analyte is a biological macromolecule, in particular a polypeptide, a polynucleotide, a polysaccharide, or a fragment of any of the above. For example, the analyte is a small molecule compound, i.e. a compound with a molecular weight of up to 1000u (1kDa). For example, the analyte is a compound that is metabolized by the body of a subject, in particular a human subject, or is a compound that is administered to a subject to induce a change in the subject's metabolism. Thus, for example, the test substance may be a drug of abuse or a metabolite thereof, such as amphetamine, cocaine, methadone, ethyl glucuronide, ethyl sulfate, an opiate, in particular buprenorphine, 6-monoacylmorphine, codeine, dihydrocodeine, morphine, morphine-3-glucuronide, and / or tramadol, and / or an opioid, in particular acetylfentanyl, carfentanyl, fentanyl, hydrocodone, norfentanyl, oxycodone, and / or oxymorphone.

[0017] For example, the test substance may be a therapeutic agent, such as valproic acid, clonazepam, methotrexate, voriconazole, mycophenolic acid (total), mycophenolic acid-glucuronide, acetaminophen, salicylic acid, theophylline, digoxin, immunosuppressants, particularly cyclosporine, everolimus, sirolimus, and / or tacrolimus, analgesics, particularly meperidine, normeperidine, tramadol, and / or O-desmethyl-tramadol, anticancer drugs, such as cyclosporine, everolimus, sirolimus, and / or tacrolimus ... biological agents, in particular gentamicin, tobramycin, amikacin, vancomycin, piperacillin (tazobactam), meropenem, and / or linezolid; antiepileptic drugs, in particular phenytoin, valporic acid, free phenytoin, free valproic acid, levetiracetam, carbamazepine, carbamazepine-10,11-epoxide, phenobarbital, primidone, gabapentin, zonisamide, lamotrigine, and / or topiramate. For example, the test substance is a hormone, in particular cortisol, estradiol, progesterone, testosterone, 17-hydroxyprogesterone, aldosterone, dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEA-S), dihydrotestosterone, and / or cortisone, e.g., the sample is a serum or plasma sample and the test substance is cortisol, DHEA-S, estradiol, progesterone, testosterone, 17-hydroxyprogesterone, aldosterone, DHEA, dihydrotestosterone, and / or cortisone, e.g., the sample is a saliva sample and the test substance is cortisol, estradiol, progesterone, testosterone, 17-hydroxyprogesterone, androstenedione, and / or cortisone, e.g., the sample is a urine sample and the test substance is cortisol, aldosterone, and / or cortisone.For example, the test substance is a vitamin, such as vitamin D, in particular ergocalciferol (vitamin D2) and / or cholecalciferol (vitamin D3), or a derivative thereof, such as 25-hydroxy-vitamin-D2, 25-hydroxy-vitamin-D3, 24,25-dihydroxy-vitamin-D2, 24,25-dihydroxy-vitamin-D3, 1,25-dihydroxy-vitamin-D2, and / or 1,25-dihydroxy-vitamin-D3. For example, the test substance is a metabolic product of the subject.

[0018] As used herein, the term "sample", also referred to as "test sample", relates to any type of composition of matter, and thus the term may refer to any sample, such as, but not limited to, a biological sample. For example, the sample is a liquid sample, such as an aqueous sample. For example, the test sample is selected from the group consisting of physiological fluids, including whole blood, serum, plasma, saliva, ocular lens fluid, tears, cerebrospinal fluid, sweat, urine, milk, peritoneal fluid, mucus, synovial fluid, peritoneal fluid, and amniotic fluid, lavage fluids, tissues, cells, and the like. However, the sample may also be a natural or industrial liquid, in particular surface or ground water, sewage, industrial wastewater, process fluids, soil leachates, and the like. For example, the sample contains or is suspected to contain at least one compound of interest, i.e., the chemical to be determined, referred to as the "analyte". The sample may contain one or more additional compounds that are not the compound to be determined, which are generally referred to as matrices as described herein above. The samples may be used directly as obtained from their respective origin or may be subjected to one or more pretreatment and / or sample preparation steps. Thus, the samples may be pretreated by physical and / or chemical methods, such as centrifugation, filtration, mixing, homogenization, chromatography, precipitation, dilution, concentration, contact with binding agents and / or detection reagents, and / or any other method deemed appropriate by the skilled artisan. One or more internal standards may be added to the sample in the sample preparation step, i.e. before, during, and / or after the sample preparation step. The sample may be spiked with an internal standard. For example, the internal standard may be added to the sample at a predefined concentration. The internal standard may be selected to be easily identifiable under normal operating conditions of the selected detector, such as, for example, a photometric cell in a mass spectrometer, for example, a UV-Vis spectrometry device, an evaporative light scattering refractometer, a conductivity meter, or any device deemed appropriate by the skilled artisan. The concentration of the internal standard may be predetermined and may be significantly higher than the concentration of the analyte.

[0019] As used herein, the term "lifetime" of a chromatography column is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically, but not exclusively, refer to a parameter that represents the wear incurred by a chromatography column due to past separations performed thereon. For example, lifetime is a parameter that indicates the remaining useful life (RUL), i.e., how many separations are expected to be possible using a chromatography column before the performance of the column becomes unacceptable. For example, lifetime is a parameter that indicates the spent life, i.e., the number of separations already performed using a chromatography column. Thus, in the case of remaining lifetime, lifetime may be expressed as the number of runs remaining, and in the case of spent lifetime, it may be the cumulative number of runs. However, it is also envisaged that lifetime is an abstract value, for example, lifetime may be expressed as a calculated percentage of initial performance, or as a lifespan score, e.g., in any unit, or any other parameter deemed appropriate by the skilled artisan. The lifetime of a chromatography column may be a column-specific parameter, as will be understood by those skilled in the art. Furthermore, the lifetime of a chromatography column may be, for example, a protocol-specific, e.g., assay-specific, parameter, i.e., for example, different protocols, particularly assays, are different in how demanding the column is in terms of performance, and thus the lifetime value may be different for different protocols and / or assays. Thus, a chromatography column may be at the end of its remaining lifetime for a demanding assay, but may still be usable for a less demanding assay.

[0020] Determining the lifetime may include predicting the lifetime in a subsequent chromatographic separation, such as the lifetime in the next 5, preferably the next 10, more preferably the next 35 chromatographic separations. The predicted lifetime corresponding to 35, 65, 130, or even 265 chromatographic separations may depend on the number of streams in the system and / or the length of the chromatographic method. For example, a chromatographic separation in a chromatographic column may have a length of 108 seconds. In this example, 33.33 chromatographic separations may be performed per hour in the chromatographic column. However, as an example, 100 injections may be performed per hour in a three-stream system.

[0021] The method may include operating a chromatography column. The term "operating a chromatography column" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a specific or special meaning. The term may specifically refer to, but is not limited to, using a chromatography column one or more times for a chromatographic separation. For example, the term may relate to the use of a chromatography column for a series of chromatographic separations, which may be separations according to the same protocol or different protocols. The term "chromatographic protocol", also called "protocol", refers to the totality of chromatographic parameters applied to a chromatography column, i.e., in particular a specific mobile phase or its gradient, temperature, pressure, flow rate, and sample type. The term "assay", as used herein, refers to the totality of parameters that define the chromatography column used and the analysis to be performed, in particular the analyte to be determined, as well as the protocol, further including sample preparation steps, such as those specified elsewhere herein. Thus, in a particular chromatographic column, it is in principle possible to detect several non-identical analytes using the same protocol, i.e. to use the same protocol for two or more non-identical assays. However, it is also possible to detect the same analyte with different protocols. As is clear from the above, the use of different protocols to detect the same analyte, as well as the use of the same protocol to detect different analytes, defines in each case a specific assay. In contrast, the term "separation", which may also be called "run" or "injection" or "measurement", relates to a single event of carrying out chromatography using a particular chromatographic column, independent of the protocol and / or assay, for example. Nevertheless, a separation is typically carried out using one specific protocol and in the context of a specific assay.

[0022] The method includes, by way of example, the following steps, which may be performed in the given order. It should be noted, however, that different orders are also possible. Furthermore, it is also possible to perform one or more of the method steps once or repeatedly. Furthermore, it is possible to perform two or more method steps simultaneously or overlapping in time. The method may include further method steps not listed.

[0023] The method comprises: i) receiving model input chromatographic data via at least one communication interface; ii) determining, using at least one processing device, at least one data-driven model based on the model input chromatographic data, at least one state variable representative of the life of the chromatographic column; iii) determining information about the lifetime by evaluating the determined state variables using a processing device; Including, The evaluation includes comparing the determined state variable to at least one threshold value.

[0024] Method steps i)-iii) may be performed fully automatically, specifically using a processing device. As used herein, the term "processing device" is a broad term and should be given its ordinary and customary meaning to those skilled in the art, and should not be limited to a specific or special meaning. The term may specifically, but not exclusively, refer to any logic circuitry configured to perform basic operations of a computer or system, and / or may generally refer to a device configured to perform calculations or logical operations. In particular, the processing device may be configured to process basic instructions that drive the computer or system. By way of example, the processing device may include at least one arithmetic logic unit (ALU), at least one floating point unit (FPU), such as a numeric coprocessor or numeric coprocessor, a number of registers, specifically registers configured to supply operands to the ALU and store the results of the operations, and memories, such as L1 and L2 cache memories. In particular, the processing device may be a multi-core processor. In particular, the processing unit may be or comprise a central processing unit (CPU) or a graphics processing unit (GPU), or a tensor processing unit (TPU). Additionally or alternatively, the processing unit may be or comprise a microprocessor, and thus in particular the elements of the processing unit may be included in one single integrated circuit (IC) chip. Additionally or alternatively, the processing unit may be or comprise one or more application specific integrated circuits (ASICs) and / or one or more field programmable gate arrays (FPGAs), etc.

[0025] The method can be automated in a fully automated MS-based analyzer, thus avoiding measurement errors and automatically reducing system downtime and costs.

[0026] As used herein, the term "communication interface" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically, but not limited to, refer to an item or element forming a boundary configured to transfer information. In particular, the communication interface may be configured to transfer information from a computing device, such as a computer, for example, to transmit or output information to another device. Additionally or alternatively, the communication interface may be configured to transfer information to a computing device, such as a computer, for example, to receive information. The communication interface may specifically provide a means for transferring or exchanging information. In particular, the communication interface may provide a data transfer connection, such as Bluetooth, NFC, inductive coupling, etc. By way of example, the communication interface may be or may comprise at least one port comprising one or more of a network or Internet port, a USB port, and a disk drive. The communication interface may be at least one web interface.

[0027] As used herein, the term "chromatographic data" is a broad term and should be given its ordinary and accustomed meaning to one of ordinary skill in the art and should not be limited to any specific or special meaning. The term may specifically refer to, but is not limited to, data determined by running a chromatographic column and / or readbacks such as printouts.

[0028] As used herein, the term "model input chromatographic data" is a broad term and should be given its ordinary and customary meaning to one of ordinary skill in the art and should not be limited to a special or special meaning. The term may specifically refer to, but is not limited to, input data for a data-driven model. The model input chromatographic data may include at least one any feature derived from at least one input variable and / or at least one state variable. For example, the model input chromatographic data may include at least one feature derived from at least one input variable selected from the group consisting of standard deviation, mean absolute deviation, median, quantile, kurtosis, maximum, minimum, autocorrelation coefficient, linear trend, fast Fourier coefficient, number of peaks, and the like. For example, the state variable may be one or more of the following: run time of a chromatographic column, volume of solvent passed through a chromatographic column, operating temperature of a chromatographic column, sample matrix, such as urine, whole blood, spinal fluid, flow rate of HPLC. For example, the model input chromatographic data may include at least one input parameter selected from the group consisting of maximum pressure, pressure difference at the beginning and end of the chromatogram, peak width, retention time, peak symmetry, sample type, assay type for each measurement, time of use of a chromatographic column in the chromatographic device, at least one maintenance parameter, e.g., related to column protection, changes in the chromatographic column. For example, the model input chromatographic data may include at least one input parameter of maximum pressure, peak width, and retention time.

[0029] The model input chromatographic data includes a plurality of inputs. The model input chromatographic data may include a plurality of features derived from a plurality of input variables and / or state variables. For example, the model input chromatographic data may include a plurality of features derived from input variables selected from the group consisting of standard deviation, mean absolute deviation, median, quantile, kurtosis, maximum, minimum, autocorrelation coefficient, linear trend, fast Fourier coefficient, number of peaks, and the like. For example, the state variables may be selected from the run time of the chromatographic column, the volume of solvent passed through the chromatographic column, the operating temperature of the chromatographic column, the sample matrix, e.g., urine, whole blood, spinal fluid, HPLC flow rate. For example, the model input chromatographic data may include a plurality of input parameters selected from the group consisting of maximum pressure, pressure difference at the start and end of the chromatogram, peak width, retention time, peak symmetry, sample type, assay type for each measurement, time of use of the chromatographic column in the chromatographic device, at least one maintenance parameter, e.g., related to column protection, and changes in the chromatographic column. For example, the model input chromatographic data may include maximum pressure, peak width, and retention time as input parameters. In contrast to known methods, such as those described in U.S. Pat. No. 5,670,379, the model input chromatographic data may include multiple inputs. This can allow for a significant improvement in prediction. U.S. Pat. No. 5,670,379 uses only a single input, in particular pressure or peak width or retention time or lamp intensity or detector noise / drift. However, such an approach may be problematic in terms of robustness to noise.

[0030] The term "sample type" as used herein includes any parameter that affects the type and amount of sample components. For example, a sample type is defined at least by the sample matrix and the pre-purification state of this sample. The term "sample matrix" is known to relate to the totality of the components other than the analyte of a sample, and the sample matrix is ​​defined, for example, by the origin of the sample, for example, a body fluid sample such as whole blood, serum, plasma, urine, saliva or sputum, or a tissue sample such as a biopsy. The term "pre-purification state" of a sample applies to the sample after it has been obtained and relates to the totality of treatments that at least partially remove sample components, in particular matrix components. Pre-purification steps are known in the art and include, for example, centrifugation, precipitation, solvent treatment, extraction, homogenization, heat treatment, freezing and thawing, cell lysis, application to a pre-column, etc., among others, as specified elsewhere herein. As understood from above, as used herein, any differences in pre-purification steps that result in differences in sample components, for example, are considered to result in different sample types; thus, for example, a serum sample centrifuged at low speed and a serum sample that is ultracentrifuged may be different sample types.

[0031] The model input chromatography data may include metadata relating to one or more of at least one chromatography column manufacturing factor, at least one laboratory specific factor. The metadata may relate to any data of the new chromatography column provided by the manufacturer, for example via a barcode, RFID, or other data carrier. For example, the metadata may include one or more of the column dimensions, for example, length, width, diameter, such as inner and / or outer diameter, internal volume, inner surface, lot information, manufacturer, and installation time.

[0032] As used herein, the term "receive model input chromatographic data" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically refer to, but is not limited to, one or more of receiving, downloading, accessing, determining, measuring, detecting, and recording the model input chromatographic data. For example, the model input chromatographic data may be obtained by downloading and / or accessing the model input chromatographic data from at least one database, such as a detector or a cloud. For example, the method may include a measurement using chromatography in step i). Specifically, the model input chromatographic data may be obtained by performing at least one chromatographic run.

[0033] The method may include receiving raw chromatographic data via a communication interface. For example, the method may include reading raw chromatographic data provided by a chromatographic device or another data source. The method may include at least one data preparation step to prepare the raw chromatographic data for model input. The data preparation step may include applying at least one data pre-processing step to the raw chromatographic data. The pre-processing step may include smoothing and / or one or more of basic data transformations such as normalization, standardization, logarithmic transformation, etc.

[0034] The model input chromatographic data may include at least one feature. The method may include at least one feature extraction or derivation step. The feature extraction or derivation step may include extracting or deriving at least one feature from raw chromatographic data preprocessed for the model input. The feature extraction or derivation step may include generating features for the model input, for example, by determining derivatives, label coding, etc. As outlined in more detail below, the machine learning model includes at least one neural network, for example at least one convolutional neural network (CNN). In the case of a CNN, feature generation based on the input may be a learning objective of the model. For example, a convolutional layer in a neural network may receive a set of values ​​as input, apply a filter function, and output a derived value that is used as an input in the next layer of the neural network. The filter function may be a determination of a linear trend in a set of input values. However, the parameters of the filter (and therefore its particular function) are generally learned by the CNN. Typically, a CNN contains not just one filter function, but several so that different features can be derived from the input.

[0035] The method may include determining whether the received model input chromatographic data includes an outlier. For example, the model input chromatographic data may be pressure data for at least one injection. The outlier detection may include determining whether the pressure progression includes at least one data point that differs significantly, for example by more than 10%, from other observations. For example, the outlier detection includes determining deviations from the pressure curve and / or abnormalities in the pressure curve, also referred to as abnormal behavior of the pressure curve. The outlier may be caused by distortions not related to column aging, for example, damage to capillaries. The outlier detection may include using at least one trained data-driven model. The trained data-driven model may use a probabilistic supervised machine learning framework designed for regression and classification tasks. For example, the trained data-driven model may be a Gaussian regression model. The data-driven model may be designed for outlier detection using at least one Gaussian regression feature. For example, outliers may be hidden in the raw pressure curve, but may be visible in the scaled pressure curve and may be detectable through at least one Gaussian regression feature. Various scaling methods may be possible, such as standardization, min / max scaling, etc. The Gaussian regression feature may be able to capture anomalous behavior in the pressure curve. This may allow the method to enable simple outlier detection in the column aging signal. The Gaussian regression feature may be used to track local changes in the data, such as anomalous behavior at the beginning of each injection, which may allow better error analysis. The outlier detection may be performed almost entirely data-driven and does not impose any limiting conditions on the data. This may allow the use of automatic feature generation for continuous system monitoring of other parts of the chromatographic apparatus. If an outlier is detected, the method may include removing and / or imputing the outlier. For example, the method may include flagging the outlier if the Gaussian regression feature used exceeds at least one threshold.Known methods, such as those described in U.S. Pat. No. 5,670,379, do not describe outlier removal. Alternatively, other outlier procedures may be implemented, where appropriate. The outlier procedure may calculate one or more distance indices for possible outliers to the remaining normal observations, or may score the deviation from the expected distribution for normal observations. For example, a large deviation of an observation from a trend smoother (such as loess) associated with a threshold may be considered an outlier. At least one clustering algorithm and / or at least one dimensionality reduction procedure (such as principal component analysis or autoencoder) may be applied. Identified outliers may be removed from the analysis or imputed by smoothed values.

[0036] Step ii) may include feeding the model input chromatographic data to the data-driven model and calculating a lifetime prediction. As used herein, the term "prediction" refers to a predicted value of a future state variable. The result of the determination in step ii) may be a predicted time series of the state variable, such as one or more of a histogram, a point estimate such as a mean and / or a median, and / or an uncertainty range such as a confidence interval, that represents the time course of the state variable.

[0037] As used herein, the term "state variable indicative of the life of a chromatography column" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a specific or special meaning. The term may specifically refer to, but is not limited to, anything that characterizes the life of a chromatography column and / or allows conclusions to be drawn regarding the life of a chromatography column. For example, the state variable is at least one variable selected from the group consisting of maximum pressure, peak width, retention time, peak symmetry.

[0038] As used herein, the term "data-driven model" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically refer to, but is not limited to, an empirical predictive model. The data-driven model may include one or more of at least one linear model, at least one nonlinear model, and at least one machine learning model. The machine learning model includes at least one neural network. For example, the data-driven model may be a physical model based on and / or using differential equations to determine aging, for example. For example, the machine learning model may include at least one recurrent neural network, such as at least one convolutional neural network (CNN), such as at least one long-short-term memory (LSTM), at least one gated recurrent unit (GRU), and / or at least one long-term recurrent convolutional network (LRCN), and at least one fractional polynomial model. The data-driven model may include at least one time series model, such as Holt-Winters triple exponential smoothing, autoregressive integrated moving average (ARIMA), etc. The data-driven model may be a feature-based model, where a feature or a subset of features is used to predict lifespan. The training data may be used to select features for the trained model. Feature selection may include selecting a subset of relevant features, particularly variables and predictors, for use in building the model. Methods for the interpretation of artificial intelligence models are known to those skilled in the art.

[0039] The data-driven model may be derived from the analysis of experimental data. The data-driven model may include at least one trained model. As used herein, the term "trained model" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a special or special meaning. The term may specifically refer to, but is not limited to, a model for predicting lifetime trained on at least one training data set, also referred to as training data. For example, the data-driven model is trained on at least one training data set. The training data set may include historical data of at least one known chromatographic column configuration. The historical data may include at least one feature, specifically, input variables such as standard deviation, mean absolute deviation, median, quantile, and / or multiple features derived from metadata. For example, the historical data includes one or more of the following operational data: pressure curve, maximum pressure, pressure difference at the beginning and end of the chromatogram, peak width, retention time, peak symmetry. The method may include generating at least one training data set and determining parameters of the data-driven model by training the data-driven model with the training data. For example, for training, available historical data may be subdivided into data used for parameter determination, i.e. data called training data set, a test data set, and a validation data set. The validation data set may be a data set used to adjust hyperparameters. The test data set may include historical data independent of the training data set. The test data set may be used to test the model trained with the training data set. Such procedures are generally known to those skilled in the art. The parameters of the model may be determined using at least one optimization algorithm. The data-driven model may be a self-learning model. The method may include updating the data-driven model taking into account the received model input chromatographic data and the determined state variables. Training the model may include continuous training, for example using input data to further optimize the model.

[0040] For example, training data may be generated by "stressing" a chromatographic column with a defined matrix until it requires replacement and measuring its performance using a defined set of test substances. Such testing may be performed in an accelerated manner, where injections may be performed at high frequency (no idle cycles) or with one or more concentrated matrices, e.g., with a higher amount of aging-related substances than would normally be expected, or in a normal operating mode. Typically, the matrix used is not random, but one defined standard matrix prepared in large quantities. The matrix may closely resemble a matrix expected in the real world. For example, training data from a customer's laboratory may be used. This can be guaranteed to be 100% realistic, but may be difficult to have during development. Additionally or alternatively, a mixed training dataset may be used that includes both example datasets.

[0041] For example, the data-driven model may be a linear model, e.g., the state variable is the pressure in the infusion, and the probability that the predicted value of the linear model exceeds a pre-specified limit is p=1-cdf((-mean+limit)*sqrt(ws) / std) is given by "p" indicates the probability that the pressure is above a prespecified limit, "mean" is the pressure predicted by the linear model, "limit" is the prespecified limit, cdf is the cumulative density function, "sqrt()" is the square root, ws is the window size (the number of data points used to determine the "mean"), and "std" is the standard deviation in the sample.

[0042] For example, instead of or in addition to a linear model, a more complex model such as a neural network may be used. For example, the neural network may be an RNN. The RNN may be designed to receive multiple input features simultaneously to calculate a pressure prediction, also called a pressure forecast. This may allow for a significant improvement in model performance. For example, the state variable may be the pressure at injection, and the model input chromatographic data may use the retention time, the pressure maximum, and the pressure difference at the start and end of the chromatogram.

[0043] For example, the data-driven model may include at least two long short-term memory (LSTM) layers. The data-driven model may include a single output node. For example, each of the LSTM layers may be designed with 25 hidden units. The window size may be fixed to 20 input values ​​and a varying number of input features. For example, training was performed using the adam optimization algorithm, a batch size of 32 samples, and a total number of epochs of 100. Five training columns with a total number of samples of 2305 were used as the training set.

[0044] For example, the data-driven model may include at least two LSTM layers. The first LSTM layer may be used as an encoder layer and the second LSTM layer may be used as a decoder layer. The first LSTM layer, designed with, for example, 25 hidden units, may be used as an encoder of the input window. The data-driven model may further include at least one attention layer, which may be designed to weight the hidden states in the encoder layer. The output of this layer may be fed to a second LSTM layer, which acts as a decoder and outputs a sequence of time steps, for example, 50 time steps. The window size may be fixed to 20 input values ​​and a variable number of input features. For example, training was performed using the adam optimization algorithm, a batch size of 32 samples, and a total number of epochs of 100. Five training columns with a total number of samples of 2305 were used as the training set.

[0045] For example, the data-driven model may be a fractional polynomial model, where the state variable is the pressure at injection, and the model input chromatographic data includes the retention time, the pressure maximum, and the pressure difference at the start and end of the chromatogram.

number

[0046] Usually, a person skilled in the art would use a simple Taylor series, as described for example in US Patent No. 5,670,379, whereas the present invention proposes to use one or more of the above mentioned models.

[0047] Step iii) involves determining information regarding the lifetime by evaluating the determined state variables by using a processing device. If step ii) involves determining a plurality of state variables, step iii) may involve evaluating a set of state variables. For example, pressure and hold time may be used, for example, with individual thresholds. As used herein, the term "evaluate" is a broad term and should be given its ordinary customary meaning to those skilled in the art and should not be limited to a special or particular meaning. Specifically, without limitation, this term may refer to applying at least one mathematical operation, for example, at least one comparison, to the determined state variables. Evaluation includes comparing the determined state variables with at least one threshold. As used herein, the term "threshold" is a broad term and should be given its ordinary customary meaning to those skilled in the art and should not be limited to a special or particular meaning. Specifically, without limitation, this term may refer to at least one predetermined value or at least one predetermined range of state variables that are assumed to relate to a lifetime value that ensures that the chromatographic column is still suitable for a given assay. The threshold may be selected to ensure that the chromatographic column meets at least one applicable quality criterion. For example, with respect to pressure, the maximum pressure of the HPLC pump may be used, for example, with an additional safety margin. The threshold may depend on the manufacturer of the pump. For example, in the case of Agilent's Infinity II, the threshold may be about 1000 bar. For example, the threshold may be ±2.5% for the hold time. For example, in the case of FWHM, at least one derived quality criterion such as resolution: ≧1.25 and / or tailing factor: <2 may be used. For example, in the case of FWHM, at least one percentage deviation from the target value, for example, ±2%, may also be used. For example, step iii) may include comparing the prediction determined in step ii) with at least one threshold and calculating the probability of column failure.For example, the lifetime information may include one or more of a probability of failure, a pressure at a future time, a remaining useful life, a remaining useful count, binary information such as a column change or column retention, etc. The lifetime information may include at least one output recommendation for the chromatography column.

[0048] The method may include step iv) of providing information about the lifespan via at least one user interface and / or initializing at least one maintenance process if it is determined in step iii) that the value of the state variable exceeds a threshold value. As used herein, the term "user interface" is a broad term and should be given its ordinary and customary meaning to a person skilled in the art and should not be limited to a special or special meaning. The term may refer to an element or device configured to interact with a surrounding environment, such as for the purpose of exchanging information unidirectionally or bidirectionally, for example to exchange one or more data or commands, but is not limited thereto. For example, the user interface may be configured to share information with a user and receive information by a user. The user interface may be a function for visually interacting with a user, such as a display, or a function for acoustically interacting with a user. The user interface may comprise, by way of example, one or more of a graphical user interface, a data interface, such as a wireless and / or wired data interface. Initializing the at least one maintenance process may include initializing at least one operation for handling the chromatography column and / or handling the sample to be analyzed. For example, an action may include rerouting a sample to an alternative stream in a multi-stream LC / MS.

[0049] For example, the user interface may include at least one graphical user interface (GUI) configured to display information regarding the lifetime. For example, the GUI may display instructions for manually replacing the chromatography column. The user follows the instructions on the GUI to manually replace the chromatography column. The LC system may automatically prepare the chromatography column, for example, by equilibration cycling with an appropriate eluent, and return to operation.

[0050] The proposed method may enable cost reduction. Prediction of chromatography column lifetime may reduce the cost of the assay without sacrificing performance by ensuring the most efficient use of the chromatography column. The proposed method may enable optimized throughput. System downtime may be minimized by advance warning to the operator enabling planned maintenance. The proposed method may enable improved reliability. Performance prediction allows rerouting of samples in multi-stream LC / MS, thus avoiding measurement on the wrong column and thus avoiding loss of material. The proposed method may enable improved performance. The proposed method may allow lifetime over hours to be determined with a high relative accuracy of 5-10%. The proposed data-driven prediction may take into account individual laboratory and column manufacturing factors and may be superior to classical statistical methods. The proposed method may enable improved patient safety. Performance prediction may be automated in fully automated MS-based analyzers, thus avoiding significant measurement errors in the IVD environment.

[0051] In a further aspect, a test system is proposed, configured to perform the method for determining the life span according to the present invention. As used herein, the term "system" is a broad term and should be given its ordinary and customary meaning to those skilled in the art and should not be limited to a specific or special meaning. The term may specifically, but not exclusively, refer to any set of interacting or interdependent component parts that form a whole. In particular, the components may interact with each other to perform at least one common function. At least two components may be treated independently or may be coupled or connectable.

[0052] This test system is - at least one communication interface configured to receive model input chromatographic data; - at least one processing device configured to determine at least one state variable representative of a lifetime of the chromatography column using at least one data-driven model based on the model input chromatography data, the processing device configured to determine information regarding the lifetime by evaluating the determined state variable, the evaluation comprising comparing the determined state variable to at least one threshold value; and - at least one user interface configured to provide information regarding the service life and / or to initiate at least one maintenance process; Equipped with.

[0053] With regard to the embodiments and definitions of the test system, reference is made to the embodiments and definitions of the method for determining a lifetime described above or given in more detail below.

[0054] In a further aspect, a computer program for determining the lifetime of at least one chromatography column of at least one chromatography apparatus is proposed, which computer program, when executed on a computer or a computer network, is configured to cause the computer or the computer network to execute the method for determining the lifetime of at least one chromatography column of at least one chromatography apparatus according to the invention, and is configured to execute at least steps i) to iii) and optionally also step iv) of the method for determining the lifetime. In particular, the computer program may be stored on a computer-readable data carrier and / or a computer-readable storage medium.

[0055] As used herein, the terms "computer-readable data carrier" and "computer-readable storage medium" may specifically refer to non-transitory data storage means such as a hardware storage medium having computer-executable instructions stored thereon. A computer-readable data carrier or storage medium may specifically be or comprise a storage medium such as a random access memory (RAM) and / or a read-only memory (ROM).

[0056] Thus, in particular, one, two or more or all of the method steps a) and b) and optionally step iv) as described above may be carried out using a computer or a computer network, preferably using a computer program.

[0057] Further disclosed and proposed herein is a computer program product having program code means for executing the method according to the invention in one or more of the embodiments contained herein when the program is executed on a computer or a computer network. In particular, the program code means may be stored on a computer readable data carrier and / or a computer readable storage medium.

[0058] Further disclosed and proposed herein is a data carrier having stored thereon a data structure which, after being loaded into a computer or computer network, for example into a working or main memory of the computer or computer network, is capable of performing a method according to one or more of the embodiments disclosed herein.

[0059] Further disclosed and proposed herein is a computer program product having program code means stored on a machine-readable carrier for executing the method according to one or more of the embodiments disclosed herein when the program is executed on a computer or computer network. As used herein, a computer program product refers to a program as a tradeable product. The product may generally be in any format, such as a paper format, or on a computer-readable data carrier and / or a computer-readable storage medium. In particular, the computer program product may be distributed via a data network.

[0060] Finally, a modulated data signal containing instructions readable by a computer system or computer network for carrying out a method according to one or more of the embodiments disclosed herein is disclosed and proposed herein.

[0061] With reference to computer-implemented aspects of the present invention, one or more or all of the method steps of the method according to one or more of the embodiments disclosed herein may be performed by using a computer or a computer network.Thus, in general, any of the method steps including providing and / or manipulating data may be performed by using a computer or a computer network.In general, these method steps may include any method steps, except for those method steps that typically require manual operations, such as providing a sample and / or performing the actual measurement in certain aspects.

[0062] Specifically, in this specification, a computer or computer network comprising at least one processor, the processor being configured to execute a method according to one of the embodiments described herein, - a computer-loadable data structure configured, when executed on a computer, to carry out a method according to one of the embodiments described in this specification; - a computer program configured, when it is run on a computer, to carry out a method according to one of the embodiments described in this specification, a computer program comprising program means for carrying out a method according to one of the embodiments described herein when said computer program is run on a computer or on a computer network, a computer program comprising program means according to the preceding embodiment, the program means being stored on a computer readable storage medium; - a storage medium storing a data structure, the data structure being configured to execute a method according to one of the embodiments described herein after being loaded into a main memory and / or a working memory of a computer or a computer network, and - a computer program product comprising program code means storable or stored on a storage medium, which, when executed on a computer or on a computer network, performs a method according to one of the embodiments described herein. is further disclosed.

[0063] In a further aspect, a method for operating a chromatography column is proposed. The method includes, by way of example, the following steps, which may be performed in a given order. It should be noted, however, that a different order is also possible. Furthermore, it is also possible to perform one or more of the method steps once or repeatedly. Furthermore, it is possible to perform two or more method steps simultaneously or overlapping in time. The method may include further method steps not listed.

[0064] The method comprises: (a) performing a plurality of chromatographic separations of a sample in the chromatography column; (b) providing a model input chromatographic data relating to at least a portion of said chromatographic separation; and (c) determining the lifespan of said chromatography column according to a method for determining lifespan according to the present invention. Includes.

[0065] For example, step a) comprises applying at least one column void volume, in a further example at least one column volume, of sample and mobile phase to said chromatography column. This step may further comprise applying additional mobile phase, a mobile phase gradient, and / or applying a re-equilibration step to the chromatography column. This step may also comprise detecting one or more analytes after separation by means known to those skilled in the art, and / or collecting one or more fractions for further analysis. This step may further comprise performing mass spectrometry on at least a portion of the eluate from the chromatography column.

[0066] For example, the chromatography column may be removed from service or modified if the determined lifetime exceeds a threshold, such as if the lifetime falls outside a predefined reference range or exceeds a threshold.

[0067] In summary, without excluding further embodiments, the following embodiments can be envisaged: Embodiment 1. A computer-implemented method for determining the lifetime of at least one chromatography column of at least one chromatography apparatus, comprising: i) receiving model input chromatographic data via at least one communication interface; ii) determining, using at least one processing device, at least one data-driven model based on the model input chromatographic data, at least one state variable representative of the life of the chromatographic column; iii) determining information about the lifetime by evaluating the determined state variables using a processing device; Including, The method, wherein the evaluating includes comparing the determined state variable to at least one threshold value.

[0068] Embodiment 2. The method of the preceding embodiment, wherein determining the lifetime includes predicting the lifetime in a subsequent chromatographic separation, such as the lifetime in the next 5, preferably the next 10, more preferably the next 35 chromatographic separations.

[0069] Embodiment 3. The method of any one of the preceding embodiments, wherein the model input chromatographic data includes at least one input parameter selected from the group consisting of maximum pressure, pressure difference at the start and end of the chromatogram, peak width, retention time, peak symmetry, sample type, assay type for each measurement, time of use of the chromatographic column on the chromatographic device, at least one maintenance parameter, and changes in the chromatographic column.

[0070] Embodiment 4. The method of any one of the preceding embodiments, wherein the model input chromatographic data includes metadata regarding one or more of: at least one chromatographic column manufacturing factor; at least one laboratory specific factor.

[0071] Embodiment 5. The method of any one of the preceding embodiments, wherein the state variable is at least one variable selected from the group consisting of maximum pressure, pressure difference at the start and end of the chromatogram, peak width, retention time, peak symmetry.

[0072] Embodiment 6. The method of any one of the preceding embodiments, wherein the data-driven model includes one or more of at least one linear model, at least one nonlinear model, at least one machine learning model, at least one time series model, and the machine learning model includes at least one recurrent neural network, such as at least one long short-term memory (LSTM), at least one gated recurrent unit (GRU), and / or at least one convolutional neural network (CNN), such as at least one long-term recurrent convolutional network (LRCN), and at least one fractional polynomial model.

[0073] Embodiment 7. The data-driven model is a linear model, and the probability that the predicted value of the state variable exceeds a pre-specified limit is p=1-cdf((-mean+limit)*sqrt(ws) / std) is given by The method of embodiment 6, wherein "p" indicates the probability that the pressure is above a pre-specified limit, "mean" is the pressure predicted by the linear model, "limit" is the pre-specified limit, cdf is the cumulative density function, "sqrt()" is the square root, ws is the window size, and "std" is the standard deviation in the sample.

[0074] Embodiment 8. The method according to embodiment 6, wherein the data-driven model is a recurrent neural network, the state variable is the pressure at injection, and the retention time, maximum pressure, and pressure difference at the start and end of the chromatogram are used as model input chromatographic data, the data-driven model includes at least two long-short-term memory (LSTM) layers, the first LSTM layer is used as an encoder layer, and the second LSTM layer is used as a decoder layer, and the data-driven model further includes at least one attention layer, the attention layer is designed to weight the hidden state in the first LSTM layer, and the output of the attention layer is fed to the second LSTM layer, which outputs a sequence of time steps.

[0075] Embodiment 9. The data-driven model is a fractional polynomial model, the state variable is the pressure at injection, and the model input chromatographic data are the retention time, the maximum pressure, and the pressure difference at the start and end of the chromatogram, and the fractional polynomial model is:

number

[0076] Embodiment 10. The method of any one of the preceding embodiments, wherein the data-driven model is trained on at least one training data set, the training data set including historical data of at least one known chromatographic column configuration, the historical data including one or more of the following operational data: pressure curve, maximum pressure, pressure difference at the beginning and end of a chromatogram, peak width, retention time, peak symmetry.

[0077] Embodiment 11. The method according to the preceding embodiment, comprising generating at least one training data set and determining parameters of the data-driven model by training the data-driven model on the training data, the parameters being determined using at least one optimization algorithm.

[0078] Embodiment 12. The method of any one of the preceding embodiments, wherein the data-driven model is a self-learning model, and the method includes updating the data-driven model taking into account the received model input chromatographic data and the determined state variables.

[0079] Embodiment 13. The method of any one of the preceding embodiments, wherein the information regarding the life span includes one or more of the following: probability of failure, pressure at a future time, and remaining useful life.

[0080] Embodiment 14. The method of any one of the preceding embodiments, wherein the information regarding the lifetime includes at least one output recommendation for the chromatography column.

[0081] Embodiment 15. The method of any one of the preceding embodiments, further comprising step iv) of providing information regarding the service life via at least one user interface and / or initializing at least one maintenance process if it is determined in step iii) that the value of the state variable exceeds a threshold value.

[0082] Embodiment 16. The method according to the preceding embodiment, wherein initializing at least one maintenance process includes initializing at least one operation for handling a chromatography column and / or handling a sample to be analyzed.

[0083] Embodiment 17. The method of any one of the preceding two embodiments, wherein the user interface comprises at least one graphical user interface (GUI) configured to display information regarding the life span.

[0084] Embodiment 18. A method according to any one of the preceding embodiments, comprising receiving raw chromatographic data via a communications interface, and including at least one data preparation step for preparing the raw chromatographic data for model input, the data preparation step including applying at least one data pre-processing step to the raw chromatographic data, the pre-processing step including one or more of: smoothing, basic data transformation.

[0085] Embodiment 19. The method according to the preceding embodiment, wherein the model input chromatographic data includes at least one feature, and the method includes at least one feature extraction or derivation step, the feature extraction or derivation step including extracting or deriving at least one feature from raw chromatographic data that has been preprocessed for model input.

[0086] Embodiment 20. A test system configured to carry out the method according to any one of the preceding embodiments, comprising: - at least one communication interface configured to receive model input chromatographic data; - at least one processing device configured to determine at least one state variable representative of a lifetime of the chromatography column using at least one data-driven model based on the model input chromatography data, the processing device configured to determine information regarding the lifetime by evaluating the determined state variable, the evaluation comprising comparing the determined state variable to at least one threshold value; and - at least one user interface configured to provide information regarding the service life and / or to initiate at least one maintenance process; A test system comprising:

[0087] Embodiment 21. A computer program for determining the life of at least one chromatography column of at least one chromatography apparatus, the computer program being configured, when executed on a computer or a computer network, to cause the computer or the computer network to perform a method for determining the life of at least one chromatography column of at least one chromatography apparatus according to any one of the preceding embodiments referring to the method for determining the life of at least one chromatography column, the computer program being configured to perform at least steps i) to iii) and optionally iv) of the method for determining the life of at least one chromatography column of at least one chromatography apparatus according to any one of the preceding embodiments referring to the method for determining the life of at least one chromatography column.

[0088] Embodiment 22. A method for operating a chromatography column, comprising: (a) performing a plurality of chromatographic separations of a sample in the chromatography column; (b) providing model input chromatographic data relating to at least a portion of said chromatographic separation; (c) determining the lifetime of the chromatography column according to any one of the methods of embodiments 1 to 19; The method includes:

[0089] Embodiment 23 The method of embodiment 21, wherein use of the chromatography column is discontinued or altered if the lifetime exceeds a threshold value.

[0090] Further optional features and embodiments are disclosed in more detail in the following description of the embodiments, preferably in conjunction with the dependent claims, where each optional feature may be realized in an independent manner or in any possible combination, as understood by a person skilled in the art. The scope of the present invention is not limited by the preferred embodiments. The embodiments are illustrated diagrammatically in the figures, where the same reference numbers in these figures refer to the same or functionally equivalent elements. [Brief description of the drawings]

[0091] [Figure 1] 1 shows an exemplary embodiment of a chromatography device and test system in a schematic diagram. [Diagram 2] 1 shows a diagram of exemplary maximum pressure values ​​obtained from a number of chromatographic separations. [Diagram 3] 1 illustrates a flow diagram of an exemplary embodiment of a computer-implemented method for determining the life of at least one chromatography column of at least one chromatography apparatus. [Figure 4] 1 illustrates a flow diagram of an exemplary embodiment of a computer-implemented method for determining the life of at least one chromatography column of at least one chromatography apparatus. [Diagram 5] 1 illustrates a flow diagram of an exemplary embodiment of a computer-implemented method for determining the life of at least one chromatography column of at least one chromatography apparatus. [Figure 6A] 1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 6B] 1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 6C] 1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 6D]1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 6E] 1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 6F] 1 shows an example result diagram of a method for determining life span with a data-driven model comprising a linear model. [Figure 7A] FIG. 6B shows relative error plots for the results of FIGS. 6A-6F. [Figure 7B] FIG. 6B shows relative error plots for the results of FIGS. 6A-6F. [Figure 7C] FIG. 6B shows relative error plots for the results of FIGS. 6A-6F. [Figure 8A] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8B] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8C] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8D] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8E] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8F] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8G] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 8H] 1 shows an illustration of an exemplary result of a method for determining longevity with a data-driven model comprising a first machine learning model. [Figure 9A] 13 shows an example result diagram of a method for determining longevity with a data-driven model comprising a second machine learning model. [Figure 9B] 13 shows an example result diagram of a method for determining longevity with a data-driven model comprising a second machine learning model. [Figure 9C] 13 shows an example result diagram of a method for determining longevity with a data-driven model comprising a second machine learning model. [Figure 9D] 13 shows an example result diagram of a method for determining longevity with a data-driven model comprising a second machine learning model. [Figure 10A] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 10B] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 10C] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 10D] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 10E] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 10F] 1 shows an example result diagram of information regarding the lifetime of a chromatography column obtained by carrying out a method for determining the lifetime. [Figure 11] 1 shows a flow diagram of an exemplary embodiment of a method for operating a chromatography column. [Figure 12A] The experimental results are shown. [Figure 12B] The experimental results are shown. [Figure 12C] The experimental results are shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0092] Detailed Description of the Preferred Embodiments FIG. 1 shows, in a schematic diagram, an exemplary embodiment of a chromatography device 110 and a test system 112. The chromatography device 110 may include, for example, at least one liquid chromatography device 114. The liquid chromatography device 114 may be or include at least one high performance liquid chromatography (HPLC) device or at least one micro liquid chromatography (μLC) device. The chromatography device 110 may include at least one chromatography column 116. In the example shown in FIG. 1, the chromatography device 110 may be a single column device. However, other embodiments are possible, such as a multi-column device having multiple columns 116. The chromatography column 116 may have a stationary phase through which a mobile phase is pumped to effect separation and / or elution and / or migration of analytes of interest. 1, the chromatography column 116 may include at least one interface 118 for introducing a mobile phase into the stationary phase. The interface 118 for introducing the mobile phase may include one or more of a pump, a solvent reservoir, a mixing vessel, and / or a valve, etc.

[0093] The chromatography device 110 may be coupled to a mass analyzer 122, for example, via at least one interface 120. The mass analyzer 122 may be or include at least one quadrupole mass analyzer 124. The interface 120 coupling the chromatography device 110 and the mass analyzer 122 may include at least one ionization source 126 configured to generate molecular ions and transfer the molecular ions to a gas phase.

[0094] The chromatography device 110 may be connected to a test system 112, in particular via at least one communication interface 128, in particular for transferring chromatography data. The test system 112 is configured to perform a method 140 for determining a lifetime according to the present invention, such as a method according to any one of the embodiments disclosed above and / or any one of the embodiments disclosed in more detail below. An exemplary embodiment of the method 140 is illustrated in Figures 3-5.

[0095] The test system 112 includes: at least one communication interface 128 configured to receive model input chromatographic data; - at least one processing unit 130 configured to determine at least one state variable representative of a lifetime of the chromatography column 116 using at least one data-driven model based on the model input chromatography data, the processing unit 130 configured to determine information regarding the lifetime by evaluating the determined state variable, the evaluation comprising comparing the determined state variable to at least one threshold value; at least one user interface 132 configured to provide information regarding the service life and / or to initiate at least one maintenance process; Equipped with.

[0096] The user interface 132 may, for example, comprise at least one graphical user interface (GUI) configured to display information regarding the lifetime. For example, the GUI may display instructions for manually replacing the chromatography column 116. The user follows the instructions on the GUI to manually replace the chromatography column 116. The LC system may automatically prepare the chromatography column 116, for example, by equilibration cycling with an appropriate eluent, and return to operation.

[0097] As shown in FIG. 1, the communications interface 128, the processing unit 130, and the user interface 132 of the test system 112 may be connected to one another, particularly for the purposes of unidirectional or bidirectional data exchange.

[0098] FIG. 2 shows an example of chromatographic data 134, particularly a diagram of maximum pressure values ​​derived from a number of chromatographic separations. In this example, the chromatographic data 134 may include an injection pressure 136 as an input parameter. The injection pressure 136 may be equivalent to the maximum pressure in the chromatographic column 116. Thus, the injection pressure 136 may be used equivalently to the maximum pressure in the chromatographic column 116 as an input parameter. The injection pressure 136 is shown in the diagram of FIG. 2 as a function of the injection number 138. The chromatographic data 134 may be used, particularly, as input data for a data-driven model, and thus may be model input chromatographic data, as outlined in more detail below. As seen in FIG. 2, the injection pressure 136 increases as the injection number 138 increases. In general, pressure rise may be a common effect in aging of chromatography columns, and therefore the pressure at injection 136 may provide a good basis for a state variable representative of the life of the chromatography column 116.

[0099] However, other options are possible for the model input chromatographic data, such as model input chromatographic data including at least one input parameter selected from the group consisting of maximum pressure, pressure difference at the start and end of the chromatogram, peak width, retention time, peak symmetry, sample type, assay type for each measurement, time of use of the chromatographic column 116 in the chromatographic device 110, at least one maintenance parameter, e.g., related to column protection, changes in the chromatographic column 116.

[0100] 3 shows a flow diagram of an exemplary embodiment of a computer-implemented method 140 for determining the life of at least one chromatography column 116 of at least one chromatography device 110. The method 140 includes the following steps, which may be performed in a given order, by way of example. It should be noted, however, that different orders are possible. Furthermore, it is also possible to perform one or more of the method steps once or repeatedly. Furthermore, it is possible to perform two or more method steps simultaneously or overlapping in time. The method 140 may include further method steps not listed.

[0101] Method 140 is i) receiving model input chromatographic data via at least one communication interface 128 (denoted by reference numeral 142); ii) determining, using at least one processing device 130, at least one data-driven model based on the model input chromatographic data, at least one state variable representative of the life of the chromatographic column 116 (depicted by reference numeral 144); iii) determining lifetime information by evaluating the determined state variables using the processing unit 130 (depicted by reference numeral 146); Including, The evaluation includes comparing the determined state variable to at least one threshold value.

[0102] Determining the lifetime may include predicting the lifetime in subsequent chromatographic separations, such as the lifetime in the next 5, preferably the next 10, more preferably the next 35 chromatographic separations. For example, 35, 65, 130, 265 chromatographic separations may relate to 1 hour, 2 hours, 4 hours, and 8 hours. For example, determining the lifetime may include predicting the lifetime in the next 35 chromatographic separations. This allows time to react depending on the result of the lifetime determination. However, the number of chromatographic separations may be related to the design of the system, and may depend in particular on the number of streams. For a 3-stream system, 35 injections may be used. For a 6-stream system, for example, 17 injections may be used.

[0103] The method 140 may include a step iv) (indicated by reference numeral 148) of providing information regarding the lifetime via the at least one user interface 132 and / or initializing at least one maintenance process if it is determined in step iii) that the value of the state variable exceeds a threshold value. Initializing the at least one maintenance process may include initializing at least one operation for handling the chromatography column 116 and / or handling the sample to be analyzed. For example, the operation may include rerouting a sample to an alternative stream in a multi-stream LC / MS.

[0104] The lifetime information may include one or more of the following: probability of failure, pressure at a future time, remaining useful life, remaining useful count, binary information such as column change or column retention, etc. The lifetime information may include at least one output recommendation for the chromatography column 116.

[0105] The data-driven model in step ii) of method 140 may include one or more of at least one linear model, at least one non-linear model, at least one machine learning model, at least one time series model. Figure 4 illustrates an exemplary embodiment of a computer-implemented method 140 for determining the lifetime of at least one chromatography column 116 of at least one chromatography device 110, where in the example of Figure 4, the data-driven model comprises a linear model.

[0106] The method 140 may include receiving (indicated by reference numeral 150) raw chromatography data 134 via the communications interface 128. For example, the method 140 may include reading the raw chromatography data 134 provided by the chromatography device 110 or another data source. As shown in Figure 4, the sub-steps of receiving the raw chromatography data 134 and reading the raw chromatography data 134 may form part of method step i) (indicated by reference numeral 142).

[0107] 4, method step ii) (indicated by reference number 144) may include multiple different sub-steps. Method 140 may include at least one data preparation step (indicated by reference number 152) for preparing raw chromatographic data 134 for model input. Data preparation step 152 may include applying at least one data pre-processing step to raw chromatographic data 134. The pre-processing step may include smoothing and / or one or more of basic data transformations such as normalization, standardization, logarithmic transformation, etc. For example, a Savitzky-Golay filter may be applied to raw chromatographic data 134.

[0108] The model input chromatographic data may include at least one feature. The method 140 may include at least one feature extraction or derivation step (indicated by reference number 154). The feature extraction or derivation step 154 ​​may include extracting or deriving at least one feature from the raw chromatographic data 134 preprocessed for the model input. The feature extraction or derivation step 154 ​​may include generating features for the model input, for example, by determining derivatives, label coding, etc. In the example of FIG. 4, the feature extraction or derivation step 154 ​​may include one or more partial steps, in particular one or more of a partial step of determining a derivative (indicated by reference number 156), a partial step of determining a linear prediction (indicated by reference number 158), and a partial step of determining a "mean" and / or a standard deviation (indicated by reference number 160).

[0109] Further, in the exemplary embodiment of Figure 4, step ii) may include feeding the model input chromatographic data to a data-driven model and calculating a lifetime prediction (indicated by reference numeral 162). The result of the determination in step ii) may be a predicted time series of the state variables representing the time progression of the state variables, such as one or more of a histogram, a point estimate such as a mean and / or a median, and / or an uncertainty range, e.g., a confidence interval.

[0110] For example, the state variable is the pressure in the injection 136 and the probability that the linear model prediction exceeds a pre-specified limit is: p=1-cdf((-mean+limit)*sqrt(ws) / std) is given by "p" indicates the probability that the pressure is above a prespecified limit, "mean" is the pressure predicted by the linear model, "limit" is the prespecified limit, cdf is the cumulative density function, "sqrt()" is the square root, ws is the window size (the number of data points used to determine the "mean"), and "std" is the standard deviation in the sample.

[0111] As outlined above, the subsequent method step iii) (indicated by reference number 146) comprises evaluating the determined state variables and comparing the determined state variables to at least one threshold value. For example, step iii) may comprise comparing the prediction determined in step ii) to at least one threshold value and calculating a probability of failure of the chromatography column 116. Step iv) (indicated by reference number 148) may comprise providing information regarding the lifetime via at least one user interface 132. As indicated by arrow 164 in FIG. 4, the method 140 may be repeated starting from step i), for example, if the prediction falls below the threshold value and / or if the information regarding the lifetime indicates that the chromatography column 116 is still suitable for a given assay.

[0112] 5 shows a flow diagram of another embodiment of a computer-implemented method 140 for determining the life of at least one chromatography column 116 of at least one chromatography device 110. The method 140 may begin at reference numeral 166. In a subsequent method step, indicated by reference numeral 168, a variable indicating the number of injections may be set to zero and metadata may be obtained.

[0113] The model input chromatography data may include metadata regarding one or more of at least one chromatography column manufacturing factor, at least one laboratory specific factor. The metadata may relate to any data of the new chromatography column 116 provided by the manufacturer, for example, via a barcode, RFID, or other data carrier. For example, the metadata may include one or more of the column dimensions, for example, length, width, diameter, such as inner and / or outer diameter, internal volume, inner surface, lot information, manufacturer, and installation time.

[0114] A subsequent method step, indicated by reference number 170, may include increasing the variable indicative of the number of injections by a predefined increment, specifically an increment of 1. The following steps may include step i) (indicated by reference number 142), as outlined in more detail above. Method step i) may be followed by a decision node, indicated by reference number 172, which includes determining whether the received model input chromatographic data includes an outlier. If so, the method 140 may proceed to method step 174, which may include removing and / or imputing the outlier. If the model input chromatographic data does not include an outlier, the method 140 may proceed to a further decision node 176. Known methods, such as those described in US Pat. No. 5,670,379, for example, do not describe removing outliers. Decision node 176 may include determining whether the model input chromatographic data exceeds a threshold or height level. If the model input chromatographic data is below the threshold or height level, method 140 may return to method step 170. If the model input chromatographic data is above the threshold or height level, method 140 may proceed to method step ii) (indicated by reference numeral 144) and method step iii) (indicated by reference numeral 146), as outlined in more detail above.

[0115] In the example of FIG. 5, the method 140 may include a decision node 178 following method steps ii) and iii). At the decision node 178, it may be determined whether the determined state variable exceeds a threshold value. If the determined state variable exceeds the threshold value, the method 140 may proceed to method step iv) (indicated by reference numeral 148), which may indicate in particular that the use of the chromatography column 116 must be discontinued or modified. In this case, the method 140 may stop at reference numeral 180. If the determined state variable is below the threshold value, the method 140 may proceed to method steps 170, 142, 172, and 174 described above. However, if at the decision node 172, it is determined that the model input chromatographic data does not include an outlier, the method 140 may proceed to a further decision node 182. The decision node 182 may include determining whether a variable indicative of the number of injections exceeds a predetermined threshold value, for example a predetermined threshold value of 5000 injections. If the variable indicative of the number of injections exceeds a predetermined threshold, method 140 may proceed to method step iv) (indicated by reference numeral 148), in particular indicating the end of life of chromatography column 116, and method 140 may stop at reference numeral 180. If the variable indicative of the number of injections falls below a predetermined threshold, method 140 may return to method step ii) (indicated by reference numeral 144) and method step iii) (indicated by reference numeral 146).

[0116] 6A-6F show diagrams of exemplary results of a method 140 for determining a lifetime 140 with a data-driven model comprising a linear model. The results of FIGS. 6A-6F may be obtained when performing the method 140 according to the invention, for example according to the embodiment described with reference to FIG. 4. However, other embodiments are also feasible, such as any one of the embodiments described with reference to FIGS. 3 and 5. In the example shown in FIGS. 6A-6F, the state variable is the pressure at injection 136. In FIGS. 6A-6C, the pressure at injection 136 is shown as a function of the number of injections 138. Here, the predicted pressure at injection 184, the average predicted pressure at injection 186 and the observed pressure at injection 188 are shown in a combined diagram. In FIGS. 6D-6F, the probability of failure 190 is shown as a function of the number of injections 138. In particular, the probability of failure 190 may be determined according to the formula specified above, as described with reference to FIG. 4. Additionally, in Figures 6A and 6D, a total of 65 data points were used to determine the state variables, while in Figures 6B and 6E, a total of 130 data points were used, and in Figures 6C and 6F, a total of 265 data points were used.

[0117] The quality of the predictions shown in Figures 6A-6F may be evaluated when referring to Figures 7A-7C. Figures 7A-7C show plots of relative error 192 for the results of Figures 6A-6F. Specifically, the relative error 192 may be obtained by comparing the average of predicted pressure 186 to the observed pressure 188 as a function of injection number 138. Figure 7A shows the relative error 192 for the prediction of Figure 6A, Figure 7B shows the relative error 192 for the prediction of Figure 6B, and Figure 7C shows the relative error 192 for the prediction of Figure 6C. Focusing on the ±10% relative error 192 as shown by the dashed line 194 in Figures 7A-7C, the majority of the relative error 192 may be within the ±10% range. Only a small fraction of the relative error 192 may exceed ±10%, but still be less than ±15%.

[0118] Figures 8A-8H show exemplary results of method 140 for determining lifetime with a data-driven model comprising a first machine learning model. The results shown in Figures 8A-8H may be obtained when executing method 140 according to the present invention according to any of the exemplary embodiments described with reference to, for example, Figures 3-5.

[0119] In the example of Figures 8A-8H, a more complex model such as a neural network may be used instead of or in addition to the linear model. For example, the neural network may be a recurrent neural network (RNN). The RNN may be designed to receive multiple input features simultaneously to calculate a pressure prediction, also called a pressure forecast. This can make it possible to significantly improve model performance. For example, the state variable is the pressure 136 in the injection, and as model input chromatographic data, the retention time, the maximum pressure value, and the pressure difference at the start and end of the chromatogram may be used.

[0120] Figures 8A-8D show the pressure 136 in the injection as a function of the number of injections 138. Here, the average predicted pressure 186 in the injection and the pressure 188 observed in the injection are shown in the combined figure. Figures 8E-8H show plots of the relative error 192 for the results of Figures 8A-8D. Specifically, the relative error 192 can be obtained by comparing the average predicted pressure 186 with the observed pressure 188 as a function of the number of injections 138. The relative error 192 in Figure 8E corresponds to the prediction in Figure 8A, the relative error 192 in Figure 8F corresponds to the prediction in Figure 8B, the relative error 192 in Figure 8G corresponds to the prediction in Figure 8C, and the relative error 192 in Figure 8H corresponds to the prediction in Figure 8D. Further, Figures 8A and 8E are shown for training with a total of 50 epochs, Figures 8B and 8F are shown for training with a total of 100 epochs, Figures 8C and 8G are shown for training with a total of 200 epochs, and Figures 8D and 8H are shown for training with a total of 300 epochs.

[0121] In the examples of FIGS. 8A-8H, the data-driven model is trained with at least one training data set. The training data set may include historical data of at least one known chromatographic column configuration. The historical data may include at least one feature, specifically, input variables such as standard deviation, mean absolute deviation, median, quantiles, and / or a plurality of features derived from metadata. For example, the historical data may include one or more of the following operational data: pressure curve, maximum pressure, pressure difference at the beginning and end of the chromatogram, peak width, retention time, peak symmetry. The method 140 may include generating at least one training data set and determining parameters of the data-driven model by training the data-driven model with the training data. For example, for training, the available historical data may be subdivided into data used for parameter determination, i.e., data referred to as a training data set, a test data set, and a validation data set. The validation data set may be a data set used to tune hyperparameters. The test data set may include historical data independent of the training data set. The test data set may be used to test the model trained with the training data set. Such procedures are generally known to those skilled in the art. The parameters may be determined using at least one optimization algorithm. The data-driven model may be a self-learning model. The method 140 may include updating the data-driven model taking into account the received model input chromatographic data and the determined state variables. Training the model may include, for example, continuous training using the input data to further optimize the model.

[0122] For example, the data-driven model may include at least two long short-term memory (LSTM) layers. The data-driven model may include a single output node. For example, each of the LSTM layers may be designed with 25 hidden units. The window size may be fixed at 20 input values ​​and a variable number of input features. For example, training was performed using the adam optimization algorithm, a batch size of 32 samples, and a total of 50 to 300 epochs. Five training columns with a total number of samples of 2305 were used as the training set. As can be seen in Figures 8A to 8H, in order to avoid overfitting the machine learning model, training may be stopped at a preferably low total number of epochs, specifically less than 200 epochs, more specifically 100 epochs.

[0123] 9A-9D show diagrams of exemplary results of a method 140 for determining lifespan with a data-driven model comprising a second machine learning model. The results shown in Figs. 9A-9D may be obtained when performing the method 140 according to the present invention, for example according to any of the exemplary embodiments described with reference to Figs. 3-5.

[0124] In the example of Figures 9A-9D, as an alternative to the machine learning model described with reference to Figures 8A-8H, the data-driven model may comprise at least two LSTM layers. The first LSTM layer may be used as an encoder layer and the second LSTM layer may be used as a decoder layer. The first LSTM layer, designed with, for example, 25 hidden units, may be used as an encoder of the input window. The data-driven model may further comprise at least one attention layer, which may be designed to weight the hidden states in the encoder layer. The output of this layer may be fed to a second LSTM layer, which acts as a decoder and outputs a sequence of time steps, for example consisting of 50 time steps. The window size may be fixed to 20 input values ​​and a different number of input features. For example, training was performed using the adam optimization algorithm, a batch size of 32 samples, and a total of 50 (Figures 9A and 9C) and 100 (Figures 9B and 9D) epochs. Five training columns with a total number of samples of 2305 were used as the training set.

[0125] 9A and 9B again show the pressure 136 at injection as a function of the number of injections 138. Here, the average predicted pressure 186 at injection and the observed pressure 188 at injection are shown in a combined diagram. In FIGS. 9C and 9D, a diagram of the relative error 192 for the results of FIGS. 9A and 9B is shown, where the relative error 192 in FIG. 9C corresponds to the prediction in FIG. 9A and the relative error 192 in FIG. 9D corresponds to the prediction in FIG. 9B. The relative error 192 may be obtained by comparing the average predicted pressure 186 with the observed pressure 188 as a function of the number of injections 138. As can be seen in FIGS. 9C and 9D, the relative error 192 is within the range of ±10%.

[0126] 10A-10F are diagrams showing exemplary results of information regarding the lifetime of a chromatography column 116 obtained by executing a method 140 for determining the lifetime. The results shown in Figs. 10A-10F may be obtained when executing the method 140 according to the present invention according to any one of the exemplary embodiments described with reference to Figs. 3-5 or according to any other possible embodiment, in particular with a data-driven model comprising at least one machine learning model.

[0127] In Figures 10A and 10B, the pressure at injection 136 is shown as a function of the number of injections 138. Specifically, the observed pressure at injection 188 is shown in these figures. Figures 10C and 10D show the predicted remaining useful life 196 as a function of the number of injections 138. Figures 10E and 10F show the relative error 192, where the relative error 192 in Figure 10E corresponds to the prediction in Figure 10C and the relative error 192 in Figure 10F corresponds to the prediction in Figure 10D. The relative error 192 may be obtained by comparing the average 186 of the predicted pressure with the observed pressure 188 as a function of the number of injections 138. As can be seen in Figures 10E and 10F, the relative error 192 is within the range of ±10%.

[0128] 11 shows a flow diagram of an exemplary embodiment of a method for operating a chromatography column 116. The method includes the following steps, which may be performed in the given order by way of example. It should be noted, however, that different orders are also possible. Furthermore, it is also possible to perform one or more of the method steps once or repeatedly. Furthermore, it is possible to perform two or more method steps simultaneously or overlapping in time. The method may include further method steps not listed.

[0129] The method comprises: (a) performing a plurality of chromatographic separations of a sample in the chromatography column 116 (indicated by reference numeral 198); (b) providing model input chromatographic data relating to at least a portion of said chromatographic separation (shown generally at 200); and (c) determining the lifespan of said chromatography column 116 according to a method for determining lifespan 140 according to the present invention (indicated by reference numeral 140); Includes.

[0130] For example, step a) includes applying at least one column void volume of sample and mobile phase, in a further example at least one column volume, to said chromatography column 116. This step may further include applying additional mobile phase, a mobile phase gradient, and / or applying a re-equilibration step to the chromatography column 116. This step may also include detecting one or more analytes after separation by means known to those skilled in the art, and / or collecting one or more fractions for further analysis. This step may further include performing mass spectrometry 122 on at least a portion of the eluate from the chromatography column 116.

[0131] For example, the chromatography column 116 may be taken out of service or modified if the determined lifetime exceeds a threshold, such as if the lifetime falls outside a predefined reference range or exceeds a threshold.

[0132] Figures 12A-12C show the experimental results for estradiol. For the high pressure liquid chromatography (LC) method used for Figures 12A-12C, the following experimental conditions were used: [Table 1]

[0133] Gaussian feature generation may be performed as follows: Finite basis function expansions can be used to approximate any function (such as a pressure curve).

number

[0134] For a single pressure curve, a Gaussian basis function may be used.

number

[0135] A grid of Gaussian functions for i = 1,...,n in the (discrete) domain [a,b] may be defined by:

number

number

[0136] The finite basis function expansion defined above can be considered as a linear regression model, so in FIG. 12, the coefficient β 0 is referred to as the "intercept", and the n basis function coefficients β 1 , , β n is simply referred to as “1”,…,“n”. The joint set of all coefficients β determined by this technique is called the “Gaussian feature”.

[0137] FIG. 12A shows the feature-scaled Gaussian basis values ​​(left plot) and pressure (in [MPa]) (right plot) for normal estradiol as a function of index. The lower table shows the intercepts. FIG. 12A shows in the lower table and left plot that a small number of Gaussian features can approximate high-dimensional pressure data, allowing sufficient reproduction of its salient characteristics (right plot).

[0138] FIG. 12B shows the feature scaled Gaussian basis values ​​(left plot) and pressure (in [MPa]) (right plot) for outlier estradiol as a function of index. The lower table shows the intercept. FIG. 12B shows that for abnormal pressure curves (right plot), the Gaussian feature values ​​change significantly compared to normal pressure curves in the preceding plots (lower table and left plot). This property can therefore be exploited for outlier detection.

[0139] 12C shows an exemplary outlier detection using Gaussian features. Normal and outlier pressure curves can be distinguished in Gaussian feature space because their feature values ​​are significantly different. This allows classification using fairly simple multivariate methods (e.g., logistic regression or random forests). [Explanation of symbols]

[0140] 110 Chromatography equipment 112 Test System 114 Liquid Chromatography Equipment 116 Chromatography Columns 118 Interface 120 Interface 122 Mass spectrometer 124 Quadrupole mass spectrometer 126 Ionization Source 128 Communication Interface 130 Processing equipment 132 User Interface 134 Chromatography Data 136 Pressure in Injection 138 injections 140 Methods for determining life expectancy 142 Receive model input chromatographic data 144 Determine at least one state variable 146 Evaluate the determined state variables 148 Provide information on life expectancy 150 raw chromatographic data received 152 Data Preparation Steps 154 Feature Extraction or Derivation Step 156 Determining Derivatives 158 Determine the Linear Prediction 160 Determine the "mean" and / or standard deviation 162 Model Input Feed the chromatographic data into the data-driven model to calculate lifetime predictions 164 Arrow 166 Start 168 Set i=0 and get metadata 170 Increasing variables 172 Decision Node 174 Removal and / or imputation of outliers 176 Decision Node 178 Decision Node 180 Stop 182 Decision Node 184 Predicted pressure for injection 186 Mean predicted pressure for injection 188 Pressure observed during injection 190 Probability of failure 192 Relative Error 194 dashed line 196 Remaining useful life 198 Performing Multiple Chromatographic Separations 200 Provide model input chromatographic data

Claims

1. 1. A computer-implemented method (140) for determining a lifespan of at least one chromatography column (116) of at least one chromatography device (110), comprising: i) receiving, via at least one communication interface (128), model input chromatographic data comprising a plurality of inputs; ii) using at least one processing device (130), determining at least one state variable representative of a life of the chromatography column (116) using at least one data-driven model based on the model input chromatography data, the data-driven model including at least one machine learning model, the machine learning model including at least one long short-term memory (LSTM), at least one recurrent neural network such as at least one gated recurrent unit (GRU), and / or at least one convolutional neural network (CNN) such as at least one long-term recurrent convolutional network (LRCN); iii) determining lifetime-related information by evaluating the determined state variables using the processing device (130), the evaluation comprising comparing the determined state variables with at least one threshold value; A computer-implemented method (140) comprising:

2. The method of claim 1 , wherein determining the lifetime comprises predicting a lifetime in a subsequent chromatographic separation.

3. 2. The method (140) of claim 1, wherein the model input chromatography data includes a plurality of input parameters selected from the group consisting of maximum pressure, pressure difference at the start and end of a chromatogram, peak width, retention time, peak symmetry, sample type, assay type for each measurement, time of use of the chromatography column (116) on the chromatography device (110), at least one maintenance parameter, and changes in the chromatography column (116).

4. 10. The method of claim 1, wherein the model input chromatography data includes metadata regarding one or more of at least one chromatography column manufacturing factor and at least one laboratory specific factor.

5. 2. The method of claim 1, wherein the state variable is at least one variable selected from the group consisting of maximum pressure, pressure difference at the beginning and end of a chromatogram, peak width, retention time, and peak symmetry.

6. 2. The method of claim 1, wherein the data-driven models further comprise one or more of at least one linear model, at least one nonlinear model, at least one time series model, and at least one fractional polynomial model.

7. 2. The method of claim 1, wherein the data-driven model is the recurrent neural network, the state variable is the pressure at injection, and the retention time, maximum pressure, and pressure difference at the start and end of the chromatogram are used as model input chromatographic data, the data-driven model includes at least two long short-term memory (LSTM) layers, a first LSTM layer is used as an encoder layer and a second LSTM layer is used as a decoder layer, and the data-driven model further includes at least one attention layer, the attention layer is designed to weight hidden states in the first LSTM layer, and an output of the attention layer is fed to the second LSTM layer, which outputs a sequence of time steps.

8. 2. The method (140) of claim 1, wherein the information regarding the lifetime includes one or more of a probability of failure (190), a pressure at a future time, and a remaining useful life (196), and / or the information regarding the lifetime includes at least one output recommendation for the chromatography column (116).

9. 2. The method (140) of claim 1, further comprising a step iv) of providing lifetime information via at least one user interface (132) and / or initializing at least one maintenance process if it is determined in step iii) that the value of the state variable exceeds the threshold value, wherein the initialization of the at least one maintenance process comprises initializing at least one operation for handling the chromatography column (116) and / or handling a sample to be analyzed.

10. A test system (112) configured to perform the method (140) of any one of claims 1 to 9, comprising: at least one communication interface (128) configured to receive model input chromatographic data comprising a plurality of inputs; at least one processing device (130) configured to determine at least one state variable representative of a lifetime of the chromatography column (116) using at least one data-driven model based on the model input chromatography data, the at least one processing device (130) configured to determine lifetime information by evaluating the determined state variable, the evaluation comprising comparing the determined state variable to at least one threshold value; at least one user interface (132) adapted to provide information regarding said lifetime and / or to initiate at least one maintenance process; A test system (112).

11. 1. A computer program for determining a service life of at least one chromatography column (116) of at least one chromatography device (110), comprising: The computer program is configured, when executed on a computer or a computer network, to cause the computer or the computer network to carry out a method (140) for determining the lifetime of at least one chromatography column (116) of at least one chromatography device (110) according to any one of claims 1 to 9, The computer program is configured to perform at least steps i) to iii) of a method (140) for determining the lifetime of at least one chromatography column (116) of at least one chromatography device (110) according to any one of claims 1 to 9, Computer program.

12. A method for operating a chromatography column (116), comprising: (a) performing a plurality of chromatographic separations of a sample in said chromatography column (116); (b) providing model input chromatographic data relating to at least a portion of said chromatographic separation; (c) determining the life of the chromatography column (116) according to a method (140) according to any one of claims 1 to 9; A method comprising:

13. 13. The method of claim 12, wherein use of the chromatography column (116) is discontinued or modified if the lifetime exceeds a threshold value.

Images