Analysis of vibrational fluorescence from biological cells

Abstract

To provide an analysis of vibration fluorescence from a biological cell.SOLUTION: A method and a system are for analyzing vibration fluorescence showing vibration ion flux associated with one or more biological cells. An exemplary method (50) may include a step (54) of detecting fluorescence (91) from one or more biological cells (72) to generate a series of data points (120) describing a vibration pattern (110). A series of gradients may be calculated (60) with respect to the vibration pattern (110). For example, sliding window (122) may be used to define a subset (126) of a series of data points (120) from which the series of gradients are calculated. The peaks (118, 150a, 150b) of the vibration pattern (110) may be identified using a series of gradients (62).SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] (Related Application) This application claims the benefit of U.S. Provisional Application No. 62 / 866524, filed on Jun. 25, 2019, the content of which is incorporated herein by reference in its entirety.

[0002] The present disclosure relates to methods and systems for analyzing oscillatory fluorescence from biological cells. More specifically, the disclosed embodiments relate to methods and systems for analyzing oscillatory fluorescence that represents oscillatory ion fluxes, such as calcium oscillations, in biological cells and for testing the effect of cell treatments on the oscillatory fluorescence.

Background Art

[0003] Introduction Most drugs fail in clinical studies due to cardiotoxicity or neurotoxicity. To reduce the occurrence of these failures, sensitive in vitro assays are required to reliably evaluate the harmful effects of compounds on heart and nerve cells before clinical studies are initiated. These in vitro assays can facilitate and rationalize drug development.

[0004] Culture systems for cardiomyocytes and neurons have been developed, and these cells exhibit spontaneous synchronized ion fluxes, such as calcium oscillations. Calcium oscillations can be detected by labeling the cells with a fluorescent calcium indicator.

[0005] Imaging systems, including analytical software, have also been developed to record and analyze vibrational fluorescence representing intracellular calcium oscillations. These imaging systems are used to measure the effects of pharmacological compounds on calcium oscillations in cultured cardiomyocytes and neurons. Significantly, these calcium oscillations are disrupted in vitro by pharmacological compounds with known cardiotoxic or neurotoxic effects. Therefore, this methodology shows promise for predicting the efficacy and dosage of drug candidates prior to safety testing and clinical studies of drugs and other chemicals. The vibrational patterns generated when calcium oscillations are disturbed by drugs and other compounds can be extremely complex. Better methods and systems are needed to detect the biologically relevant features of these complex vibrational patterns and extract values ​​for the most useful readings. [Overview of the project] [Means for solving the problem]

[0006] This disclosure provides methods and systems for recording and analyzing vibrational fluorescence representing vibrational ion flux associated with one or more biological cells. An illustrative method of analysis may include the step of detecting fluorescence from one or more biological cells to generate a set of data points describing a vibrational pattern. A set of gradients may be calculated with respect to the vibrational pattern. A sliding window may be used to define a subset of a set of data points from which a set of gradients is calculated. Peaks of the vibrational pattern may be identified using the set of gradients. Another illustrative method of analysis may include the step of detecting fluorescence from one or more biological cells to generate a set of data points describing a vibrational pattern. Primary and secondary peaks in the vibrational pattern may be identified, if present. Aspects of the secondary peaks may be determined. The present invention provides, for example, the following: (Item 1) Method of analysis (50), wherein the method is To generate a series of data points (120) describing an oscillation pattern (110), the fluorescence (91) representing oscillation ion flux associated with one or more biological cells (72) is detected (54), Calculating a series of gradients related to the vibration pattern (110) (60), Using the aforementioned series of gradients, identify the peaks (118, 150a, 150b) of the vibration pattern (110) (62) Methods that include... (Item 2) The method according to item 1, wherein the calculation (60) uses a sliding window (122) and then defines a subset (126) of the set of data points (120) from which the set of gradients is calculated. (Item 3) The method according to item 2, further comprising selecting the size of the sliding window (122) from a plurality of permitted sizes, wherein the size of the sliding window (122) corresponds to the number of data points (126) from the set of data points (120) that are encompassed by the sliding window (122). (Item 4) The size of the sliding window (122) is automatically assigned by the processor (96) based on the noise level in the vibration pattern (110) and / or the sampling interval for the set of data points (120), and the processor (96) also calculates the set of gradients and identifies the peaks (118, 150a, 150b), as described in item 3. (Item 5) The method according to item 3, wherein the size of the sliding window (122) is selected by the user and communicated to a processor (96) that similarly calculates the series of gradients and identifies the peaks (118, 150a, 150b). (Item 6) The method according to item 1, wherein the series of data points (120) are not filtered to reduce noise prior to calculating a series of gradients (60). (Item 7) The method according to item 1, wherein the peak comprises a series of primary peaks (118), the vibration pattern (110) comprises a series of events (138), each of which comprises only one of the primary peaks (118), and the method further comprises determining at least one aspect of the secondary peaks (150a, 150b) of the identified peak, each secondary peak following one of the primary peaks (118) within the event (138). (Item 8) The method according to item 7, wherein at least one aspect of the secondary peaks (150a, 150b) relates to the number, frequency, or period of the secondary peaks (150a, 150b) in the vibration pattern (110). (Item 9) The method according to item 7, wherein the vibration pattern (110) crosses a predetermined trigger level (144) twice for each event (138), and the trigger level (144) is set relative to the baseline (116) of the vibration pattern (110). (Item 10) The method according to item 1, wherein identifying peaks (118, 150a, 150b) (62) includes searching for positive-to-negative or negative-to-positive transitions in the gradients within the series of gradients. (Item 11) The method of item 10, wherein identifying a peak (62) includes filtering the peaks associated with the transition in order to obtain a set of peaks that are deemed valid. (Item 12) The method according to item 11, further comprising determining the values ​​of peak-related parameters for the set of peaks deemed to be valid. (Item 13) The method according to item 1, further comprising labeling one or more biological cells (72) with a calcium indicator (52), wherein the fluorescence (91) is emitted by the calcium indicator. (Item 14) The method according to item 1, wherein the one or more biological cells (72) include one or more cardiomyocytes or neurons. (Item 15) The method according to item 1, wherein the vibration pattern (110) comprises a series of events (138), each of which includes a single primary peak (118), and the vibration pattern (110) comprises one or more secondary peaks (150a, 150b) included in the events of the series of events (138), and the method further comprises determining (64) one or more values ​​for one or more parameters associated with the one or more secondary peaks (150a, 150b). (Item 16) A method of analysis, wherein the method is To generate a series of data points (120) describing an oscillation pattern (110), the fluorescence (91) representing oscillation ion flux associated with one or more biological cells (72) is detected (54), Identifying the primary peak (118) and secondary peaks (150a, 150b) in the vibration pattern (110) (62), To determine the side of the aforementioned secondary peaks (150a, 150b) and Methods that include... (Item 17) The method according to item 16, wherein determining the aspects of the secondary peaks (150a, 150b) includes determining the number, frequency, or period of the secondary peaks (150a, 150b). (Item 18) The method according to item 16, further comprising determining the interval regularity / irregularity of the primary peaks (118) and / or the amplitude regularity / irregularity of the primary peaks (118). (Item 19) The method of item 16, further comprising comparing the amplitude (194) of each primary peak (118) to a predetermined threshold in order to enumerate smaller peaks of the primary peak, if present. (Item 20) System (70), An optical sensor (88), wherein the optical sensor (88) is configured to detect fluorescence (91) representative of a vibrational ion flux associated with one or more biological cells (72) to generate a series of data points (120) that describe a vibrational pattern (110). A processor (96), wherein the processor (96) is configured to: (1) calculate a series of gradients with respect to the vibrational pattern (110) using a sliding window (122), and then define a subset (126) of the series of data points (120) for which the series of gradients are calculated; and (2) use the series of gradients to identify peaks (118, 150a, 150b) of the vibrational pattern (110). A system (70) comprising the above. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] [Figure 1] FIG. 1 is a flowchart of steps that may be implemented in an exemplary method for analyzing a vibrational fluorescence pattern from a biological cell.

[0008] [Figure 2] FIG. 2 is a schematic diagram of an exemplary system for detecting and analyzing a vibrational fluorescence pattern from a biological cell.

[0009] [Figure 3] FIG. 3 is a graph of a relatively simple, unperturbed vibrational pattern generated by a fluorescence signal detected from a biological cell that receives a vibrational ion flux.

[0010] [Figure 4] FIG. 4 is a graph of a series of data points for a region of the fluorescence signal of FIG. 3 that encompasses a single vibration, where the region is generally indicated at "4" in FIG. 3 and the sliding window is shown schematically at three different temporal positions, for purposes of illustrating how a subset of the data points may be selected using a sliding window for calculation of a series of gradients.

[0011] [Figure 5] Figure 5 is a graph of a single oscillation ("event") of a more complex disturbed oscillation pattern detected as shown in Figure 3, illustrating how the gradient calculated as shown in Figure 4 can be used to detect biologically relevant peaks and troughs, while ignoring smaller "spurious" fluctuations in the fluorescence signal, which are assumed to be noise.

[0012] [Figure 6] Figure 6 is a graph of only a portion of the single oscillation from Figure 5 taken around a secondary peak, illustrating how primary / secondary peaks can be filtered using predetermined amplitude and / or duration criteria to exclude small and / or transient peaks as invalid.

[0013] [Figure 7] Figure 7 is a graph of the vibration patterns detected as shown in Figure 3, illustrating aspects of the algorithm for automatically setting a baseline for the vibration patterns, i.e., setting a temporary reference line and a threshold line relative to the reference line.

[0014] [Figure 8] Figure 8 is a fragment of the lower portion of the vibration pattern in Figure 7, showing a negative peak (relative to the baseline in Figure 7) located below the threshold line and identified based on the gradient.

[0015] [Figure 9] Figure 9 is a graph of three oscillations of a complex oscillation pattern, generally detected as shown in Figure 3, where the illustrative parameters are identified.

[0016] [Figure 10]Figures 10 and 11 are graphs of the same complex vibration pattern, analyzed using different sizes of sliding windows (11 points and 5 points, respectively) where primary peaks are marked by circles and secondary peaks by diamonds. [Figure 11] Figures 10 and 11 are graphs of the same complex vibration pattern, analyzed using different sizes of sliding windows (11 points and 5 points, respectively) where primary peaks are marked by circles and secondary peaks by diamonds.

[0017] [Figure 12A] Figures 12A and 12B are excerpts of illustrative dialog boxes, representing the graphical user interface created by analysis software, with the "Options" tab of the dialog box selected. [Figure 12B] Figures 12A and 12B are excerpts of illustrative dialog boxes, representing the graphical user interface created by analysis software, with the "Options" tab of the dialog box selected.

[0018] [Figure 13A] Figures 13A and 13B are fragments of the dialog boxes shown in Figures 12A and 12B, except that the "Measurement" tab of the dialog box is selected. [Figure 13B] Figures 13A and 13B are fragments of the dialog boxes shown in Figures 12A and 12B, except that the "Measurement" tab of the dialog box is selected.

[0019] [Figure 14] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 15] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 16] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 17] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 18] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 19] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 20] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively. [Figure 21] Figures 14-21 are graphs of representative fluorescence oscillation patterns detected from cardiomyocytes treated as a control group (Figure 14) and treated with various known cardiotoxic compounds (Figures 15-21). Primary ("major") and secondary peaks are marked by software using circles and diamonds, respectively.

[0020] [Figure 22] Figure 22 is a graph plotting various peak-related readouts obtained using software such as that disclosed herein, by analyzing the fluorescence oscillation patterns detected from cardiomyocytes treated with the indicated high-risk, intermediate-risk, and low-risk TdP (torsades de pointse) compounds at indicated concentrations relative to the maximum clinical blood levels for each compound.

[0021] [Figure 23] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 24] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 25]Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 26] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 27] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 28] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 29]Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Figure 30] Figures 23-30 are graphs of representative fluorescence oscillation patterns detected from neurons treated as a control group (Figure 23) and treated with various known neurotoxic compounds (Figures 24-30). Primary and secondary peaks are marked by software using circles and diamonds, respectively. Modifications to the oscillation patterns reflect the effect of the compounds and can be characterized through multiple measurements provided by software analysis. [Modes for carrying out the invention]

[0022] Detailed explanation This disclosure provides methods and systems for recording and analyzing vibrational fluorescence representing vibrational ion flux associated with one or more biological cells. An illustrative method of analysis may include the step of detecting fluorescence from one or more biological cells to generate a set of data points describing a vibrational pattern (synonymously referred to as a vibrational trace or fluorescence trace). A set of gradients may be calculated with respect to the vibrational pattern. For example, a sliding window may be used to define a subset of a set of data points from which a set of gradients is calculated. Peaks of the vibrational pattern may be identified using the set of gradients. Another illustrative method of analysis may include the step of detecting fluorescence from one or more biological cells to generate a set of data points describing a vibrational pattern. Primary and secondary peaks in the vibrational pattern, if present, may be identified. Aspects of the secondary peaks may be determined.

[0023] The methods and systems of this disclosure may utilize a complex event analysis (CEA) algorithm designed to analyze vibrational fluorescence having several components of interest. These components may include multiple peak shapes, peak clustering, secondary peaks, regular or irregular anomalous events, irregular amplitudes or frequencies, shifting threshold levels, etc. The CEA algorithm has the ability to report multiple measurements for both primary and secondary peaks, either averaged values ​​or individual values, including many additional readouts.

[0024] The purpose of the CEA algorithm is to digitally "visualize" the general shape of biological events in order to detect and measure various parameters such as peak amplitude, rise and decay times, event duration, and frequency. To visualize biological events, non-biological transitions (noise) of spurious signals must be filtered out.

[0025] Many "point-by-point" detection methods lose accuracy due to pre-filtering that compromises data integrity or because they cannot distinguish biologically relevant transitions from noise. By the nature of its design, the CEA algorithm significantly reduces the impact of noise without pre-filtering by calculating gradients from a subset of data points that describe the vibration pattern. The gradients enable the detection of peaks and troughs while ignoring lower amplitude and / or rapid transitions (i.e., invalid peaks) caused by noise.

[0026] Further aspects of this disclosure are described in the following sections, namely (I) Definitions, (II) Method and System Overview, and (III) Examples. I. Definition

[0027] The technical terms used in this disclosure have meanings generally recognized by those skilled in the art. However, the following terms may be further defined as follows:

[0028] An event is a single vibration of an vibration pattern. An event may, among other things, begin and end at a predetermined amplitude from the baseline, and / or within a predetermined amplitude range from the baseline.

[0029] Light - ultraviolet, visible, and / or infrared radiation.

[0030] Maximum – a point in a real / conceptual graph (and / or set of points) that has a value greater than the points around it and / or is further from the baseline (e.g., lower or upper baseline) than the points around it. Multiple "maximums" are "maxima".

[0031] The minimum value – the point in a real / conceptual graph (and / or set of points) that has a smaller value than the points around it and / or is closer to the baseline (e.g., lower or upper baseline) than the points around it. Multiple "minimum" values ​​are "minima" values.

[0032] A peak is a sequence of points in an actual / conceptual graph (and / or set of points) that includes the maximum value and the points surrounding the maximum value, and optionally bounded by a pair of minimum values. A peak can be characterized according to the temporal position and amplitude of the peak's maximum value, which can collectively define the "peak position," while the amplitude of the maximum value defines the "peak value" or "peak amplitude." When the maximum value is defined relative to a baseline (e.g., a lower or upper baseline), a maximum value above the baseline forms a positive peak, while a maximum value below the baseline forms a negative peak.

[0033] Primary peak - The preceding / only peak (or valid peak) of an event.

[0034] Secondary peak - Any peak (or valid peak) following the primary peak within an event.

[0035] A trough is a sequence of points in an actual / conceptual graph (and / or a set of points) that includes the minimum value and the points around the minimum value. A trough can be characterized according to the temporal position and amplitude of the trough's minimum value, which can collectively define the "trough position," while the amplitude of the minimum value defines the "trough value" or "trough amplitude." When the minimum value is defined relative to a baseline, a minimum value above the baseline forms a positive trough, while a minimum value below the baseline forms a negative trough.

[0036] Effective peak - Any peak that meets the specified criteria for effectiveness.

[0037] A window is an algorithm for selecting a given number of data points from a dataset for processing. For example, a window with a width of 5 selects 5 points from the dataset, such as 5 consecutive points, for processing. A "sliding window" moves across the dataset, point by point, to select a subset of data points for processing. For example, a sliding window with a width of 5 points could process points 1-5, and then move point by point to process points 2-6, 3-7, and so on. II. Overview of Methods and Systems

[0038] This section provides an overview of the exemplary methods and systems of the present disclosure (see Figures 1 and 2).

[0039] Figure 1 shows a flowchart 50 of steps that may be performed in an illustrative method for analyzing vibrational fluorescence from one or more biological cells (also called “cells”). The steps may be performed in any preferred order and combination.

[0040] The cells may include any cell type that exhibits vibrational ion flux under culture conditions. Therefore, the cells may include muscle cells (i.e., cardiac muscle cells (cardiac cardiomyocytes), skeletal muscle cells, or smooth muscle cells) or nerve cells (neurons). Vibrational ion flux can occur voluntarily and spontaneously in a synchronized manner with respect to sets of cells cultured in close association with each other so that the cells can interact. Sets of cells may be associated with each other in three-dimensional assemblies such as cell spheroids, among other things, or they may be arranged as substantial monolayers. As further described below, a single set of cells may be analyzed in a container, or isolated replicate sets of cells may be exposed to different treatments in separate containers before / between analysis. Exemplary containers include petri dishes, microplate wells, flasks, or equivalents.

[0041] Cells may be obtained from any suitable source by any suitable procedure. Cells may be differentiated in vitro from stem cells. Stem cells may be, among other things, embryonic stem cells, adult stem cells, or induced pluripotent stem cells (iPSCs). In other embodiments, cells may be primary cells such as primary cardiomyocytes or primary neurons obtained from animals.

[0042] The vibrational ion flux may be an ion-specific flux or a collective ion flux. An exemplary ion-specific flux is a vibrational calcium flux that generates calcium vibrations. The flux may represent the movement of ions across the plasma membrane (i.e., into and / or out of the cell) and / or within the cell (e.g., across the membrane of the sarcoplasmic reticulum (SR) (i.e., from the SR into or out of the cytoplasm)), or equivalent.

[0043] Each set of cells may be labeled and / or treated as shown in 52. Labeling may be carried out using a fluorescent indicator having fluorescence that is sensitive to the vibrational ion flux occurring in the cells. The fluorescent indicator may be, for example, a fluorescent calcium indicator that is sensitive to intracellular calcium concentration and emits more (or less) light as the intracellular calcium concentration increases. Exemplary fluorescent calcium indicators are chemical indicators such as FLIPR® Calcium 6, FLIPR® Calcium 6-QF, Calcium Green 1, Fluo-3, Fluo-4, Fura-2, Indo-1, Oregon Green 488, Bapta-1, Fura-4F, Fura-5F, Calcium Crimson, X-rhod-1, and equivalents. Cells may be labeled with the chemical indicator by contact through the culture medium in which the cells are placed. Other exemplary fluorescent calcium indicators are genetically encoded and expressed intracellularly after the introduction of the coding sequence into the cells (e.g., via transfection, infection, or equivalents). Suitable genetically encoded calcium indicators may include Chameleon, Pericum, GCaMP, TN-L15, TN-humTnC, TN-XL, TN-XXL, Twitches, or equivalents. In other embodiments, the fluorescent indicator may be a membrane potential indicator (e.g., FluoVolt® membrane potential dye, Di-3-ANEPPDHQ, Di-4-ANEPPDHQ, etc.), a potassium indicator, a sodium indicator, a magnesium indicator, a zinc indicator, a pH indicator, or equivalents.

[0044] Each set of cells may be treated with at least one substance of interest at a selected concentration, and the substance / concentration may be tested for its effect on the cellular vibrational fluorescence detected by a fluorescent indicator, if present. The substance may be a compound such as a small molecule (e.g., a drug or drug candidate), protein, RNA, or DNA molecule, having a molecular weight of less than 10, 5, 2, or 1 kilodalton, among other things. Different compounds and / or the same compound at different concentrations may be tested on individual sets of cells to screen compounds and / or determine a dose-response curve for a given compound. Each set of cells may be treated with the substance for at least 1, 2, 5, 10, 30, or 60 minutes, or 2, 4, 6, 8, 10, 12, 18, or 24 hours, or 24–72 hours, or any preferred length of time. Treatment and labeling may be carried out in parallel, sequentially, or at overlapping times with a time offset.

[0045] Fluorescence may be detected from at least one cell of each set of biological cells, as shown in 54. This fluorescence detection samples the fluorescence signal from at least one cell, such as signal intensity over time, and generates a series of data points (also called time points) that describe an oscillation pattern with respect to the fluorescence signal. Thus, the oscillation pattern may be graphed as detected fluorescence intensity as a function of time. The fluorescence signal may be sampled at any preferred rate, generally, in particular, at a constant rate above 1 Hz, such as a rate of 1 to 100 Hz, and over any preferred sampling duration, such as at least 10, 30, or 60 seconds, or at least 2, 5, or 10 minutes. Fluorescence detection may be performed using an image sensor, an optical point sensor, or equivalent.

[0046] A baseline for the vibration pattern may be established, as shown in 56. The baseline may be set automatically using software, for example, as described below in Example 2, and / or set by a user via a graphical user interface. The baseline may be positioned below or above the vibration pattern, either by default or in response to user input (see, for example, Examples 2 and 5). In some embodiments, a temporary reference line may be set above (or below) the vibration pattern, opposite to the expected location of the baseline to be established. Peaks relative to the reference line may be identified, and linear regression of peak or trough locations may find a line that serves as a baseline during subsequent identification of peaks and troughs.

[0047] The noise level of the vibration pattern may be estimated as shown in 58. The noise level may be estimated based on signal fluctuations in the vicinity above or below the vibration pattern (see Examples 2 and 5). For example, the average of consecutive differences between the peak values ​​(i.e., maximum values) (or trough values ​​(i.e., minimum values)) in the vicinity of the baseline of the vibration pattern may be used to determine the estimate of the noise level.

[0048] A series of gradients may be calculated from a series of data points of an oscillation pattern, as shown in 60. For example, a sliding window that slides along the time axis of the oscillation pattern may be used to select a contiguous subset of data points for calculating a series of gradients by fitting a line to each subset by linear regression (see, e.g., Example 1). In other words, the sliding window may slide along the time axis of the oscillation pattern by only one point (or more points) for the calculation of each contiguous gradient. The sliding window has a fixed size that includes a fixed number of three or more points while the series of gradients are calculated with respect to a given dataset (e.g., from a given well). The size of the sliding window may be selected automatically by the software (e.g., based on the noise level of the dataset and / or the sampling rate of the data points) or selected by the user. The size of the sliding window may be changed between different datasets (e.g., each dataset is collected from a different well), which allows the size to be optimized for each dataset.

[0049] The peaks (and troughs) of each vibration pattern may be identified using gradients, as shown in 62. That is, a set of maximum and minimum values ​​may be identified based on the change in the sign of the gradient along a set of gradients (i.e., with respect to time). Each maximum value above the baseline can be closely approximated by a change in a set of gradients from positive to negative (or from negative to positive with respect to the baseline above the vibration pattern). Similarly, each minimum value above the baseline can be closely approximated by a change in a set of gradients from negative to positive (or from positive to negative with respect to the baseline above the vibration pattern). The maximum and minimum values ​​identified using a set of gradients may be refined, among other things, by local analysis of points and / or interpolation, if necessary. The maximum and / or minimum values ​​(and corresponding peaks and troughs) may also be filtered (i.e., restricted) according to one or more other predetermined criteria in order to exclude maximum (and peaks) and / or minimum values ​​(and troughs) that do not meet the predetermined criteria. The predetermined criteria may relate to the amplitude and / or duration of at least one of each corresponding peak / trough and / or event containing at least one of the peaks (see, for example, Examples 1 and 5). The predetermined criteria may be automatically assigned based on the noise level of the fluorescence signal and / or subjectively assigned by the user.

[0050] The values ​​of one or more peak-related parameters for each vibration pattern may be determined as shown in 64. The parameters and their values ​​may be synonymously referred to as read values ​​or descriptors. The values ​​may be determined using the peaks identified in 62. The values ​​may be average values ​​with respect to vibration pattern measurements (or at least with respect to some of those events) or values ​​for individual events / peaks. Exemplary values ​​may be for primary peaks only, secondary peaks only, or collectively for primary and secondary peaks. The values ​​may include individual / average primary / secondary peak amplitudes, average primary / secondary peak frequencies, average in-event frequencies of secondary peaks, event area, primary / secondary peak spacing, rise slope, decay slope, rise time, decay time, etc. Further embodiments of peak-related parameters are described in Examples 1, 3, and 5.

[0051] If present, the effect of each treatment may be assessed based on at least one of the values ​​determined from the corresponding vibration pattern, as shown in 66. For example, the effect may be assessed based on aspects of the secondary peaks in the vibration pattern, such as the number, frequency, and / or period of secondary peaks.

[0052] Any combination of the steps in Figure 1 or other methods disclosed herein may be embodied as a computer method, computer system, or computer program product. Thus, aspects of the analysis method may take the form of a purely hardware embodiment, a purely software embodiment (including firmware, resident software, microcode, and equivalents), or an embodiment combining software and hardware aspects, all of which may generally be referred to herein as “circuits,” “modules,” or “systems.” Furthermore, aspects of the analysis method may take the form of a computer program product embodied in a computer-readable medium (or multiple mediums) having computer-readable program code / instructions embodied thereon.

[0053] Any combination of computer-readable media may be used. Computer-readable media may be computer-readable signal media and / or computer-readable storage media. Computer-readable storage media may include electronic, magnetic, optical, electromagnetic, infrared, and / or semiconductor systems, apparatus, or devices, or any suitable combination thereof. More specific embodiments of computer-readable storage media include, namely, portable computer diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, any suitable combination thereof, and / or equivalents. In the context of this disclosure, computer-readable storage media may include any suitable non-transient tangible medium capable of containing or storing programs for use by or related to instruction execution systems, apparatus, or devices.

[0054] A computer-readable signal medium may include, for example, a data signal that is propagated with computer-readable program code embodied therein, either in the baseband or as part of a carrier wave. Such a propagated signal may take any of various forms, including, but not limited to, electromagnetic, optical, and / or any preferred combination thereof. A computer-readable signal medium may include any computer-readable medium, rather than a computer-readable storage medium, that is capable of communicating, propagating, or transporting a program for use by or related to an instruction execution system, apparatus, or device.

[0055] Program code, which is embodied on a computer-readable medium, may be transmitted using any suitable medium, including, but not limited to, wireless, wired, fiber optic cables, and / or any preferred combination thereof.

[0056] Computer program code for performing operations relating to aspects of the methods disclosed herein may be written in one or any combination of programming languages, including object-oriented programming languages ​​such as Java®, Smalltalk, C++, and / or equivalents, and conventional procedural programming languages ​​such as C. Mobile applications may be developed using any preferred language, including those mentioned above and Objective-C, Swift, C#, HTML5, and equivalents. The program code may run entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), and / or the connection may be made to an external computer (e.g., via the Internet using an Internet service provider).

[0057] Aspects of the methods disclosed herein are described below with reference to flowcharts and / or block diagrams of the methods, apparatus, systems, and / or computer program products. Each block and / or combination of blocks in the flowcharts and / or block diagrams may be implemented by computer program instructions. Computer program instructions may be provided to a general-purpose computer, a dedicated computer, or a processor of another programmable data processing device, and the instructions may be executed via a processor of a computer or other programmable data processing device, generating machine-readable instructions so that the instructions implement the functions / actions defined in the blocks of the flowcharts and / or block diagrams. In some embodiments, machine-readable instructions may be programmed on a programmable logic device such as a field-programmable gate array (FPGA).

[0058] Computer program instructions can also be stored in computer-readable media, which can instruct the computer to function in a particular manner, such as producing a product, including instructions that implement functions / actions defined in the blocks of a flowchart and / or block diagram.

[0059] Computer program instructions can also be loaded onto a computer to generate a computer implementation process, causing a series of operational steps on a device to be performed so that the instructions executed on the computer provide a process for implementing the functions / actions defined in the blocks of a flowchart and / or block diagram.

[0060] Any flowcharts and / or block diagrams in the drawings are intended to illustrate the architecture, functionality, and / or operation of possible implementations of the systems, methods, and computer program products disclosed herein, in terms of aspects of the methods disclosed herein. In this regard, each block may represent a module, segment, or portion of code that constitutes one or more executable instructions for implementing a defined logical function. In some implementations, the functions described in a block may be performed in an order other than that described in the drawings. For example, two blocks shown consecutively may actually be executed substantially in parallel, or blocks may sometimes be executed in reverse order, depending on the functionality involved. Each block and / or combination of blocks may be implemented by a dedicated hardware-based system (or a combination of dedicated hardware and computer instructions) that performs the defined function or action.

[0061] Figure 2 shows an illustrative system 70 for detecting and analyzing vibrational fluorescence from biological cells 72 held by a sample holder 74. The sample holder 74 is depicted here as a microplate 76 having multiple wells 78, each containing a replicated set of cells 72. The system 70 may include a stage 80 for supporting the sample holder and a drive mechanism 82 for moving the stage 80 and the sample holder 74 relative to each other and positioning each well 78 on the optical axis of the system 70.

[0062] System 70 may also include an objective lens 84, a light source 86, and an optical sensor 88 (e.g., an image sensor). The light source 86 may generate light 87 to irradiate each set of cells 72 and induce fluorescence from the cells. The light for cell irradiation may propagate to the cells via a beam splitter 90 and the objective lens 84 in the reflected illumination configuration shown in Figure 2. The light-induced fluorescence 91 may be collected by the objective lens 84, the beam splitter 90, and the objective lens 84 for propagation through an optional tube lens 92 for detection by the optical sensor 88, creating a time-dependent fluorescence signal.

[0063] Computer 94 communicates with optical sensor 88 and processes the fluorescence signal received from optical sensor 88. Computer 94 may also include a processor 96 for processing instructions, memory 98 for storing instructions, and a user interface 100 for communication between computer 94 and the user. The user interface may include a user input device such as a keyboard, mouse, or touchscreen, and a display device such as a monitor. [Examples]

[0064] Examples This section describes further embodiments and aspects of the analytical methods and systems of this disclosure. These embodiments and aspects are for illustrative purposes only and should not limit the overall scope of the invention. (Example 1) Gradient calculation

[0065] This embodiment illustrates the exemplary calculation of gradients for vibration patterns using a sliding window and the use of gradients to identify peaks and troughs in vibration patterns (see Figure 3-6).

[0066] Figure 3 shows a graph of the oscillation pattern 110, generated by a fluorescence signal 112 detected from biological cells receiving a simple oscillation flux of calcium. Fluorescence is detected from a fluorescent calcium indicator that labels the cells. Fluorescence intensity is detected at discrete time intervals, generating a series of data points, which are plotted as a function of time using point-to-point line traces.

[0067] The vibrational pattern 110 consists of a series of vibrations 114 (also called events) of increased fluorescence from a baseline 116. Each vibration 114 may have only one peak 118, as shown here, or it may have a primary peak and one or more secondary peaks, as described below. In other embodiments, the baseline 116 may be located above the vibrational pattern 110, and each event may be characterized by a fluorescence signal 112 progressing below the baseline.

[0068] Figure 4 shows a graph of a secondary set of data points 120 from the vibration pattern 110 of Figure 3 for a single vibration 114 of the fluorescence signal. The point-connection line traces in Figure 3 are omitted. A sliding window 122 may be used to select a subset of data points 120 from which, optionally, a set of best-fit lines and a corresponding set of gradients can be calculated, centered on each data point. The size of the sliding window 122 represents its width, measured parallel to the time axis. The sliding window 122 is graphically shown on the left with a solid outline and, after conceptually sliding along the time axis as indicated by the motion arrows in 124, is shown with a dashed outline at two other illustrative positions. The sliding window 122 “stops” at several incremental positions along the time axis, and then a subset 126 of data points 120 centered at each position from which gradients are calculated is selected. Here, the sliding window 122 has a width that allows for the selection of five data points 120 at each increment position, although any preferred number of three or more data points 120 may be selected. The size (i.e., duration) of the sliding window may remain constant in order to select the same number of data points 120 for the calculation of each gradient while a series of gradients are calculated with respect to a given vibration pattern 110 (i.e., a dataset of data points). The gradients may be calculated point by point. The size of the sliding window may be changed in order to recalculate a new series of gradients for the same vibration pattern or to calculate a series of gradients for a different vibration pattern. The series of gradients calculated with respect to a vibration pattern may represent a corresponding series of incremental offset positions of the sliding window 122. In other words, the sliding window 122 may be incrementally and continuously offset by fixed integer data points, such as one or two data points, and / or by fixed time increments, in particular, to select a subset 126 of data points 120. A subset 126 of data points 120 may overlap with each other if the size of the sliding window 122 exceeds the incremental offset of the sliding window with respect to consecutive subsets.

[0069] In some embodiments, the gradient is calculated at each data point, and the “sliding window” in which the gradient is calculated moves one point at a time. In these embodiments, the subset of data points selected by the window always overlap with each other. The “direction” of the gradient at each point is evaluated to assess the overall curvature of the event as implied by the change in the direction of the gradient. The gradient calculated at each point is both forward and backward, as one or more points temporally ahead of the point and one or more points temporally behind the point contribute to the gradient.

[0070] The gradient for each subset 126 may be calculated by fitting a line to the points 120 of the subset 126 using linear regression, and then taking the gradient of the line. Figure 4 shows three vectors 128 parallel to individual lines fitted to three illustrative subsets 126. The orientation of each vector 128 matches the gradient of the corresponding fitted line. Each vector 128 may be centered at the midpoint 130 of the corresponding subset 126 and may have a time component that matches the size (i.e., width) of the sliding window 122. Thus, a set of vectors 128 can be generated using the sliding window 122 to relate the size of the sliding window to a set of gradients having a corresponding set of data points 130 or positions (time and amplitude).

[0071] The series of gradients of vector 128 allows us to examine the shape of the data, particularly the direction of data deviations. With respect to the baseline located below the vibration pattern, a positive gradient reflects the rise phase, while a negative gradient is associated with the decay phase. Since each gradient is calculated from a subset 126 having three or more points 120, noise within subset 126 does not significantly affect the accuracy of the gradient if the noise level is low relative to the size of the sliding window. Thus, the noise is effectively filtered out. Changes in the direction of consecutive vectors 128 reflect the corresponding changes in the direction of transitions within the vibration pattern. Therefore, when the gradient of consecutive vectors 128 changes from positive to negative (with respect to the baseline below the vibration pattern), this indicates that the peak maximum is traversed. Similarly, at the same baseline, when the gradient changes from negative to positive, this indicates that the trough minimum is traversed. Further verification of the data points within each transition region may then be performed to more precisely determine the locations (time and amplitude) of the maximum and minimum values ​​of the vibration pattern.

[0072] The size of the sliding window 122 can be optimized for a particular dataset according to the noise level. Because the sliding window extends over a range of data points, it acts as a natural "damper" that reduces the contribution of random noise. Furthermore, the gradient does not affect the integrity of the associated data points and therefore does not alter the detection and measurement of amplitude values ​​or valid secondary peaks. Thus, the use of gradients calculated from a subset of data points selected using the sliding window enables accurate detection and measurement of various parameters of the data without the data corruption that occurs with conventional filtering.

[0073] Figure 5 shows a graph of a single oscillation 114 ("event" 138) of the more complex oscillation pattern 140 detected as in Figure 3. The baseline 116 is positioned below the oscillation pattern 140. This figure illustrates how a vector 128 with a gradient calculated as in Figure 4 may be used to search for larger peaks and troughs while ignoring smaller "spurious" fluctuations in the fluorescence signal that may result from noise instead of biologically relevant activity. (For simplification of the illustration, vector 128 is shown here with a constant total length rather than a constant time component.)

[0074] One or more predetermined criteria may be used to identify valid events 138 within the vibration pattern 140. Each criterion is satisfied by a valid event. The criteria may include at least one amplitude threshold 142 and / or at least one duration threshold to distinguish a valid event from other deviations of the fluorescence signal. For example, the amplitude threshold 142 may be a single point on a trigger level 144 (line) traversed by the fluorescence signal, and the event is characterized by a fluorescence signal traversing the trigger level in both directions. The trigger level 144 may be set at a percentage of the total amplitude span of the vibration pattern, such as 10%, 15%, or 20% of the span from the baseline 116. The trigger level may be set automatically and / or by the user, and may be constant or variable along the time axis (e.g., along the trigger line defining the trigger level). In some cases, a second amplitude threshold may be set to exclude each event and / or each peak within it that has a maximum value less than a predetermined amplitude offset from the baseline (or trigger level 144). The second amplitude threshold may be set, in particular, at a certain percentage of the total amplitude span (e.g., 25%, 50%, 60%, 70%, 80%, etc.) or as an absolute value in relative fluorescence units.

[0075] During the search along the vibration pattern 140, the start of event 138 may be detected when the fluorescence signal crosses the trigger level 144 in a direction away from the baseline 116. This crossing may initiate peak detection. Event 138 may be considered to end when the trigger level 144 is crossed in a direction toward the baseline 116. However, left and right troughs around event 138 may be marked below the trigger level 144, as described below.

[0076] The gradient of vector 128 at each point 130 may be evaluated during the search to determine the direction of the transition. (Exemplary points 130a–130h are shown here.) A positive gradient at point 130a is associated with the rise phase 146 of the fluorescence signal, and a negative gradient at point 130b is associated with the decay phase 148 of the fluorescence signal. The neighborhood of the maximum value for each peak is detected when the gradient of the consecutive vector 128 changes from positive to negative (using the baseline 116 on the lower side). For example, points 130c, 130d, and 130e of vector 128 are very close to the individual maximum values ​​for at least the primary peak 118 and secondary peaks 150a, 150b of the same event 138. The neighborhood of the minimum value for each trough is detected when the gradient of the consecutive vector 128 changes from negative to positive (using the baseline 116 on the lower side). For example, points 130f and 130g of vector 128 are very close to the individual minimums with respect to the left trough 152a and right trough 152b that boundary event 138. (The trough between peaks 118 and 150a and 150b is not clearly identified here for the sake of illustration simplicity.)

[0077] The size of the sliding window used to create vector 128 determines the magnitude of low-amplitude transitions that will be bypassed by the search. For example, transition 154 around point 130h is not identified as a peak or trough in the search because the gradient in this region does not undergo a sign change (i.e., they remain negative). The larger the window size (and therefore the length of vector 128), the larger the transition amplitudes that will be bypassed. Thus, vector 128 collectively acts as a filter that does not destroy the fluorescence signal.

[0078] Noise analysis within a narrow region surrounding each transition point can be optionally applied to further improve the detection accuracy of each peak / trough position by filtering out very short or "low-amplitude" events that may be located near better peak positions several points away.

[0079] Figure 6 shows a graph of only a portion of the single vibration 114 (and event 138) from Figure 5, taken around the secondary peak 150a. Peaks may be filtered for effectiveness according to one or more criteria, which may be related to the peak maximum value 156 and at least one adjacent left or right trough minimum value 158a, 158b. For example, a given criterion may be related to exceeding a threshold for the local amplitude 160 of each peak (measured between the maximum value 156 and the left minimum value 158a and / or the right minimum value 158b), and / or exceeding a threshold for the duration 162 of the peak measured at a given amplitude (e.g., at the left minimum value 158a). The thresholds for local amplitude 160 and / or duration 162 may be user-defined or automatically derived from the noise level of the vibration pattern and / or the sampling interval of the vibration pattern (i.e., the reciprocal of the sampling rate). (Example 2) Baseline setup and noise estimation

[0080] This embodiment illustrates an exemplary approach to automatically setting a baseline for a vibration pattern and estimating the noise level of the vibration pattern from negative peaks identified during baseline setting (see Figures 7 and 8).

[0081] The algorithm disclosed herein can automatically estimate a baseline for each dataset by measuring the amplitude of fluctuations that fall within 10% of the data amplitude span, removing outliers using interquartile range analysis, and finally fitting a line to the resulting amplitude using linear regression. The resulting regression line may be proposed or used as the baseline.

[0082] Figures 7 and 8 illustrate exemplary approaches to establishing a baseline 116 for a dataset. A temporary baseline 170 (e.g., an upper baseline) may be set. For example, the baseline 170 may be defined by pairs of points 172, 174, each representing the maximum amplitude with respect to a preceding section 176 and a succeeding section 178 of the dataset. Each section 176, 178 may, among other things, represent any preferred percentage of the dataset, such as 10, 20, 30, 40, or 50%. Points near the preceding and succeeding ends of the vibration pattern may be excluded because the ends often contain artifacts. The position of the baseline 170 may be changed at one or both ends by allowing the user to move handles 180a, 180b via a graphical user interface. Alternatively, the user may adjust the time range of the dataset considered valid for analysis at one or both ends by moving individual handles 182a, 182b along the time axis via a graphical user interface.

[0083] The amplitude span of the dataset may be calculated. The global maximum amplitude (Ymax) and global minimum amplitude (Ymin) may be found using a search. The amplitude span is defined as follows: Ymax - Ymin.

[0084] The threshold line 183 may be set toward the lower side of the amplitude span opposite the reference line 170, such as at 70%, 80%, 90%, or 95% of the amplitude span from the reference line 170. For example, the position of the threshold line 183 at 90% of the amplitude span from the reference line 170 may be calculated as follows: Ymin + 0.1 × (amplitude span). The user may adjust the threshold line 183 at one or both ends by moving individual handles 184a, 184b along the amplitude axis via a graphical user interface.

[0085] A search for peaks relative to the baseline 170 may be performed (see Figures 7 and 8). In this embodiment, the search finds negative peaks by identifying local maximum deviations (below) the baseline 170 that are similarly located below the threshold line 183. (In other words, this peak search is the reverse of the one described above with respect to Figure 5.) The search is performed using a series of gradients as described above with respect to Figures 4 and 5, the gradients may be obtained using a relatively small size with respect to a sliding window (e.g., the width of 3, 4, or 5 data points). A small sliding window provides high sensitivity for peak detection with little noise rejection. Thus, the noise level with respect to the dataset may also be estimated from the amplitude values ​​of the peaks found in the search.

[0086] Primary peaks 118 and secondary peaks 150 may be identified. Here, and in subsequent embodiments, circles mark the maximum value of primary peak 118, and diamonds mark the maximum value of secondary peak 150 (see Figure 8). These two types of peaks 118, 150 may also be marked in a distinguishable manner when the vibration pattern is graphed and displayed to the user (e.g., via a graphical user interface). However, in order to establish a baseline, all detected peaks may be recorded and treated equally, i.e., without any distinction between primary and secondary peaks.

[0087] Linear regression may be used to fit a line through the detected peak amplitude. This line may be taken as an initial approximation of the baseline 116. The trigger level 144 may initially be set at 10% of the data span above the baseline. These automatically assigned levels can subsequently be adjusted by the user via a graphical interface or by explicitly setting the values.

[0088] Prior to linear regression, data points at the maximum values ​​of peaks 118 and 150 may or may not be filtered. For example, before selecting a baseline, outliers may be removed using interquartile range analysis, etc. Points below the first quartile or above the third quartile (a 1.5 interquartile range) may be excluded.

[0089] The noise level may be estimated from peaks 118 and 150 (see Figure 8). For example, the noise estimate may be calculated as the average of the consecutive differences between the amplitudes of peaks 118 and 150 (considered as a group). The noise estimate may be used directly as the noise level, or it may be used to calculate the noise level. This noise level, alone or in conjunction with the sampling interval, can enable the automatic selection of a suitable size for a sliding window (see Example 2) with respect to a given dataset. An exemplary algorithm for this size selection is a heuristic, i.e., adding a "noise coefficient" of [noise level / 100] to [0.3 / sampling interval (in seconds)]. The algorithm may select a window size from a range of allowed sizes. The range may include only odd data points (e.g., 3, 5, 7, etc.), only even data points (e.g., 4, 6, 8, etc.), or odd and even (e.g., 3, 4, 5, 6, etc.). The noise level may also be used to automatically generate a suitable threshold value for local amplitude 160 and to reject smaller peaks as invalid (see also Examples 1 and 5). The data sampling rate may also be used to automatically generate a suitable threshold for duration 162 and to reject very short-term peaks as invalid. (Example 3) Parameters of interest

[0090] This embodiment illustrates exemplary parameters of interest that can be measured from a vibration pattern using the algorithm described herein (see Figure 9).

[0091] Three events 138 of the vibration pattern are shown. A start point 190 and an end point 192 are marked for each event 138, and the fluorescent signal crosses the trigger level 144 in the opposite direction.

[0092] Any of the parameters described herein may be reported for individual peaks / events or averaged over a series of peaks / events of a given vibration pattern. Each primary peak 118 of event 138 has a primary peak amplitude 194. The primary peak amplitude can be measured as the amplitude difference between the maximum value of the primary peak 118 and the baseline 116. A linear damping gradient 196 may be defined by a straight line extending between the maximum value of the primary peak 118 and the endpoint 192 for event 138.

[0093] A primary peak interval 198 (also called the primary peak period) is defined between the maximum values ​​of adjacent primary peaks 118. A series of events 138 defines a series of primary peak intervals 198, which may be converted into a primary peak rate for the series of events, expressed as primary peaks per unit time, such as peaks per minute (PpM).

[0094] The secondary peak period 200 is defined between the maximum values ​​of adjacent secondary peaks 150 within or between events. The secondary peak period 200 with respect to the oscillation pattern may be converted to a secondary peak rate, expressed as secondary peaks per unit time, such as peaks per minute (PpM). Each secondary peak 150 has a secondary peak amplitude 202, which may be measured between the maximum value of the secondary peak and the baseline 116.

[0095] The rise gradient 204 and the decay gradient 206 may be calculated for each event 138. These gradients may each be defined as a percentage of the maximum value of the event, such as 10-80% or 30-70% of the maximum value.

[0096] The duration 208 of each event 138 may be calculated from the maximum amplitude as a specified percentage from the maximum amplitude of the event, in particular at 10, 20, 30, 40, 50, 60, 70, 80, or 90%. Alternatively, or in addition, the duration from peak 210 may be measured for each event that begins at the maximum value of the primary peak 118 of the event. The duration from peak may be measured from the maximum value as a specified percentage from the maximum value of the event, in particular at 10, 20, 30, 40, 50, 60, 70, 80, or 90%. (Example 4) Effect of window size on peak detection

[0097] This embodiment illustrates the effect of changing the window size on the sensitivity of peak detection (see Figures 10 and 11).

[0098] The performance of a gradient-based peak-finding algorithm, calculated as in Example 1, was tested on fluorescence intensity data collected from beating cardiomyocytes. Cardiomyocytes were in vitro labeled with a fluorescent calcium indicator and treated with a cardiotoxic compound to disrupt the regular calcium oscillations present in control cardiomyocytes (see, e.g., Example 1). The same data are graphed in Figures 10 and 11, generating complex oscillation patterns 220. Each pattern consists of a series of oscillations 114, each exhibiting a longer, more irregular decay phase compared to the oscillations of the control cells. The decay phase is characterized by an early after-depolarization (EAD)-like peak (i.e., a secondary peak).

[0099] Peak detection was performed on the data using an automatically determined sliding window size of 11 points (Figure 10) or a user-selected window size of 3 points (Figure 11). The primary peak 118 of each vibration was marked with a circle, and the secondary peak 150 was marked with a diamond. In Figure 10, with a larger window size, low-amplitude transitions are ignored. In Figure 11, with a smaller window size, low-amplitude transitions are detected as peaks. Therefore, the sensitivity of peak detection can be controlled by a thoughtful selection of the window size, in conjunction with noise rejection to invalidate detected peaks that are too small / transient (see, for example, Examples 1 and 5). (Example 5) Dialog box for analysis software

[0100] This embodiment describes exemplary dialog boxes displayed by a graphical user interface to facilitate user input into the analysis software (see Figures 12A, 12B, 13A, and 13B) (see also Figure 9).

[0101] Figures 12A, 12B, 13A, and 13B show screenshots of the dialog boxes that control the list of reads (i.e., descriptors) for analysis optimization and peak analysis. The dialog boxes contain two tabs. The "Options" tab provides access to analysis settings and data ranges. The "Measurements" tab contains the nature of events that can be selected for display in the output sheet. Exemplary aspects and features of these tabs are described below. Options tab

[0102] Event polarity. Two buttons, positive and negative, are available to select event polarity (see Figure 12A). A positive event is accompanied by an increase in fluorescence above the lower baseline, while a negative event is accompanied by a decrease in fluorescence below the upper baseline.

[0103] Select the vector length. The search vector length (i.e., the size of the sliding window) may be the most important variable for good peak detection (see Figure 12A, also see Figures 4 and 5). The search vector is generated from a set of subsets of data points in the dataset, which are selected by incrementally moving the sliding window along the time axis. A search vector is generated at each point. The vector is calculated using both forward and backward data points, centered on the window width and ranging from -((window width-1) / 2 to +((window width)-1) / 2. The gradients near the opposite ends of the dataset are calculated using a appropriately decreasing number of data points. Linear regression is performed for each subset of data points to generate a set of search vectors. In other words, the sliding window determines the presence of transitions (maximum / peaks and minimum / troughs) by sliding along the dataset, checking the gradient at each data point, and noting changes in the direction of the gradient (i.e., changes in the sign of the gradient from positive to negative or negative to positive). The length of the search vector relative to the sampling rate determines the sensitivity to detecting transitions. Longer vector lengths will automatically filter out noise (i.e., events / peaks with lower amplitude and / or shorter duration), while shorter vector lengths will detect transitions with lower amplitude and shorter duration.

[0104] Three buttons, namely "Automatically Assigned Length," "Same Length for All Wells," and "Per Well," are available to select how the vector length is assigned (see Figure 12A). When "Automatically Assigned Length" is selected, the software automatically calculates the vector length using estimated sampling rates and noise amplitudes. The automatically generated vector length is generally a good starting point for data analysis. When the "Per Well" button is selected, the user can choose a different vector length for each well (i.e., for each set of data points describing each different vibration pattern).

[0105] Dynamic threshold. This option sets a lower limit for valid peaks (see Figure 12A). All detected peaks with peak amplitudes below this threshold will be considered invalid and ignored. The value of this limit is given as relative fluorescence units (RFU) above the baseline in this embodiment. Two boxes are available to input different threshold amplitude values ​​above the baseline for the left and right ends of the line defining the dynamic threshold. In other embodiments, the dynamic threshold may be set as a percentage of the amplitude span of the dataset. For example, if the data points for a given well range from 10,000 relative fluorescence units (RFU), a setting of 80% would define a cutoff limit of 8,000 RFU.

[0106] Trigger Level. This level is defined by a trigger line extending along the time axis and determines when an event is considered to have started and ended (see Figure 12A). An "event" is defined by a fluorescence signal crossing the trigger line in both directions. However, the amplitude of the event is measured relative to the baseline. Detection of left and right troughs of an event is not limited by the trigger level, and these measurements can extend to the baseline. Event duration can be measured, among other things, as the temporal distance between two lateral trough locations, or as the temporal distance between two crossing points on the trigger line relating to the event. The trigger level is set by default to 10% above the baseline (relative to the data range). This level can be set and modified individually for each dataset (well data).

[0107] Baseline level. The baseline level defines the lower end of the detection range and the position of the baseline. All amplitudes are measured relative to the baseline level, not the trough value.

[0108] Level names (dynamic threshold, trigger level, and baseline level) are color-coded in the dialog box to match the color of the corresponding line in the graph window. Each of these levels can be defined by a separate line with a left and right end. The ends can be moved separately by using the left mouse button to grab the handle at the end and move it up or down. The entire level line can be moved up or down by using the left mouse button to grab the line at any point away from the handle and move it up or down. The level line position can also be set by changing the values ​​in the separate L (left) or R (right) edit boxes, either using the spinner or by directly entering the values. The "lock" feature will move the line ends simultaneously in response to adjustments to either the right or left edge value. This feature applies only to the dialog box and does not affect mouse positioning.

[0109] Set the default levels. Pressing this button will reset the baseline and trigger levels to their default positions (see Figure 12A).

[0110] Process all wells (see Figure 12B). If this item is checked, all selected well data will be processed. If it is not checked, only the currently selected well will be processed. Dataset (see Figure 12B):

[0111] Wells. The analysis will only apply to wells that are selected and appear in the "Details" graph window. Only selected wells will appear in the graph at any given time. The well data to display can be defined by selecting individual data series in the "Well" combo box. Event exclusion group (see Figure 12B):

[0112] Minimum Amplitude. This item defines the minimum amplitude, measured from the baseline, that would be considered valid and acceptable. Deviations forming primary / secondary peaks with a maximum value below the minimum amplitude value will be rejected. The minimum amplitude is substantially identical to the “dynamic threshold” (see Figure 12A), except that the minimum amplitude is a single value instead of being defined by values ​​along a line with independently adjustable left and right ends.

[0113] Minimum Duration. This item specifies the minimum duration of an event that will be considered valid and acceptable (see Figure 12B). The event duration may be measured between left and right trough values, or between the start and end of the event, as defined by traversing the trigger level. Deviations with a duration shorter than the minimum duration value will be rejected.

[0114] Applies to all wells. If this item is checked, the amplitude and duration values ​​displayed in the edit box will apply to data from all wells. If unchecked, the values ​​specified for each individual well will be used. Noise rejection group (see Figure 12B):

[0115] Minimum Amplitude. This item defines the minimum amplitude threshold for the local amplitude of a valid peak. The local amplitude of a peak (valid or invalid) is measured as the amplitude difference between the peak's amplitude value and the amplitude value of the nearest left (and / or right) trough. Peaks with a local amplitude below the minimum amplitude threshold will be rejected as invalid.

[0116] Minimum duration. This item defines a minimum duration threshold for valid peaks. The duration for a peak (valid or invalid) is measured as the absolute value of the time difference between (a) the time value for the lower of two adjacent troughs and (b) the time value for the equivalent amplitude on the opposite side of the peak. Peaks with a duration shorter than the minimum duration threshold will be rejected as invalid.

[0117] Applies to all wells. If this item is checked, the amplitude and duration threshold values ​​that appear in the edit box will apply to data from all wells (i.e., all datasets for separate vibration patterns). If unchecked, the values ​​defined for each individual well will be used.

[0118] Automatic. The software measures the approximate noise level of each dataset. First, the peak-finding algorithm uses a short search vector length (e.g., 3) to locate all peaks ("noise peaks") within a percentage range of the amplitude span of the dataset from the baseline (e.g., 0-10%). Noise peaks may have their maximum value relative to a reference line opposite the baseline. (For example, at a lower baseline, noise peaks may be negative peaks with their maximum value relative to a reference line located above the vibration pattern (see Example 2)). The noise estimate is then calculated as the average of the absolute differences between consecutive maximum values ​​of noise peaks that fall within the interquartile range. The noise level is taken as 30% of the noise estimate. The noise estimate and noise level are only valid if the data is relatively clean and most of the noise is located near the baseline. Measurement tab (see Figures 13A and 13B):

[0119] This tab is used to specify the measurements that should be reported on the statistics page. This page displays the results for each well (group), the average of the events detected in the well, and, where applicable, the individual values ​​for each event.

[0120] Select All. This button selects all events for display (see Figure 13A).

[0121] Select none of the options. This button clears the event display selection.

[0122] Average peak amplitude. The average peak amplitude is the representative amplitude of the maximum effective primary peak of the detected event for a given well, expressed in relative fluorescence units (RFU). For calculating the representative value, each peak amplitude is measured between the maximum peak and the baseline.

[0123] The number of peaks. This descriptor is the total number of valid primary peaks that can be combined with respect to a given well, or the total number of valid primary peaks and valid secondary peaks.

[0124] Average peak rate. This descriptor represents the number of valid primary peaks (PpM) per minute. For cardiac data, this is equivalent to "heart rates per minute" (BPM).

[0125] Number of EAD-like peaks per event. This descriptor represents the number of valid secondary peaks (EAD-like peaks) in an event or across the dataset. This item reports the average number of valid secondary peaks per event, the standard deviation of the mean, and the total number of valid secondary peaks across the dataset.

[0126] Mean EAD-like peak rate (PpM). This descriptor represents the representative rate of valid secondary peaks (EAD-like peaks) detected in the dataset (for a given well), expressed as peaks per minute (PpM).

[0127] Standard deviation of the number of EAD-like peaks. The standard deviation of the number of EAD-like peaks per event.

[0128] 10-90% CTD. These descriptors represent the duration of an event in seconds at a specified level from the maximum value of the primary peak of the event to the baseline (e.g., calcium transient duration). For example, CTD10 is the duration of the event measured at an amplitude level that is 10% of the amplitude distance from the maximum value to the baseline. CTDs from 10% on the upper side to 90% on the lower side are measured. Whenever a trough is at a certain distance from the baseline, the measurement of the CTD furthest from the peak is omitted.

[0129] 10-90% CTDP. These descriptors represent the duration of an event from the temporal position of the primary peak's maximum value to a specified level from the maximum value relative to the baseline. (CTDP is the calcium transient duration from the peak.) For example, CTDP10 is the duration of an event from the peak to the baseline at a level of 10% of the amplitude distance from the peak's maximum value. 10-90% CTD is measured. Occasionally, if the trough is at a certain distance from the baseline, the measurement of the CTDP furthest from the peak is omitted.

[0130] Area. The area between the vibration trace and the baseline for a given event, measured from the start to the end of the event, as defined by the left and right troughs for the event. The unit for area is RFU·seconds.

[0131] Peak Interval. This measurement determines the regularity (in seconds) of the interval between valid primary peaks by comparing the standard deviation and mean of the intervals. The measurement reports either uniform intervals (OK) if the standard deviation is below a threshold percentage of the mean (e.g., 50%), or irregular intervals (IRREG) if the standard deviation is above the threshold percentage.

[0132] Gradient. The percentage of the gradient measurement to and from the peak is defined within the associated edit box. Typically, the gradient is measured from 10–80% or 30–70% of the peak value. (Example 6) Compound testing using cardiomyocytes

[0133] This example illustrates results obtained from testing cardiotoxic compounds against cardiomyocytes for their effects on various parameters measured by the software described in Example 5 (see Figure 14-22).

[0134] The development of biologically relevant predictive cell-based assays for compound screening and toxicity assessment is a major challenge in drug discovery. The focus of this study was to establish a high-processing-capacity-compatible cardiotoxicity assay using human induced pluripotent stem cell (iPSC)-derived cardiomyocytes. To assess the usefulness of human iPSC-derived cardiomyocytes as an in vitro arrhythmia induction model, concentration-dependency and response were evaluated for 28 drugs associated with low, intermediate, and high torsades de pointes (TdP) risk categories. (Compounds were evaluated using Comprehensive in Vitro Proarrhythmia Assay) (Proposed by the CiPA Initiative) The effects of various compounds on the contraction rate and pattern of spontaneous cardiomyocyte activity were measured by high-speed kinematic fluorescence using calcium-sensitive dyes. 2+ It was monitored by changes in vibration. Advanced image analysis methods were used. 2+ This assay was implemented to provide multi-parameter characterization of the vibration pattern. This assay allows for the characterization of parameters such as pulse frequency, amplitude, peak width, and rise and decay times. The results demonstrate the usefulness of hiPSC cardiomyocytes for detecting in vitro drug-induced proarrhythmic effects.

[0135] iPSC-derived cardiomyocytes generate spontaneous, synchronized calcium oscillations. High-speed fluorescence imaging using the FLIPR® Penta® system revealed intracellular Ca using the EarlyTox® cardiotoxicity kit. 2+ Ca in cardiomyocytes, which is monitored by changes in levels 2+ The patterns and frequencies of the vibrations were used to measure them. In addition to a set of 28 known cardiotoxic compounds, several benchmark compounds and negative control groups were tested in the assay.

[0136] iPSC-derived cardiomyocytes: Cryopreserved human iPSC-derived cards from Cellular Dynamics International (CDI) iCell® cardiomyocytes 2 were used for the experiment. The cells were thawed and plated in 384-well format plates (Corning) at 20,000 cells / well (96-well format) or 10,000 cells / well, and cultured in maintenance medium for 7 days. The presence of strong synchronous contractions in 3D culture was visually confirmed prior to performing the experiment. In addition, 384-well plates for assays containing cardiomyocytes 1 were obtained from Ncardia, Inc. The plates were shipped pre-plated and allowed to recover for 2 days after arrival.

[0137] Cardiomyocytes were exposed to the compound for 15, 30, 60, or 90 minutes, or for 24 hours.

[0138] Intracellular Ca 2+ Vibrations were assessed using EarlyTox® calcium dye (Molecular Devices) according to a standard protocol, and cells were loaded with the dye for two hours prior to measurement.

[0139] Measuring calcium oscillations in iPSC-derived cardiomyocytes is a promising method for toxicity assessment. This work focuses on evaluating a set of 28 CiPA compounds categorized as high, medium, or low risk based on clinical data.

[0140] The FLIPR® Penta® system, equipped with a new high-speed camera, enables better resolution of calcium oscillation patterns in cardiomyocytes. ScreenWorks® Peak Pro® 2 software allows for complex event analysis and detailed pattern characterization using over 20 available pattern descriptors. The assay can be used to test drugs under development and screen chemicals for potential cardiotoxicity.

[0141] Figure 14-21 shows representative traces of calcium oscillations in control and compound-treated cardiomyocytes, recorded over 2 minutes, starting after 30 minutes of treatment with the indicated compounds and concentrations. Disturbances to the calcium oscillation patterns are described for each compound.

[0142] The control group trace (DMSO-treated cardiomyocytes) is shown in Figure 14. Only the primary peak 118 (marked with a circle) is detectable. The primary peaks have uniform amplitude and spacing from one another.

[0143] Figure 15 shows traces from cardiomyocytes treated with E-4031 at a concentration of 1 μM. A primary peak 118 and a secondary peak 150 are present. The primary peak is elongated.

[0144] Figure 16 shows traces from cardiomyocytes treated with ibutylide at a concentration of 1 μM. A primary peak 118 and a secondary peak 150 are present. The primary peak is elongated.

[0145] Figure 17 shows traces from cardiomyocytes treated with dofetilide at a concentration of 10 μM. Only the primary peak 118 is detected, but it is irregularly spaced apart from one another.

[0146] Figure 18 shows traces from cardiomyocytes treated with quinidine at a concentration of 10 μM. Primary peak 118 and secondary peak 150 are detected. The primary peak is elongated and has irregular spacing from any one another.

[0147] Figure 19 shows traces from cardiomyocytes treated with sotalol at a concentration of 10 μM. Primary peak 118 and secondary peak 150 are detected. The primary peak is elongated and has irregular spacing from any one another.

[0148] Figure 20 shows traces from cardiomyocytes treated with astemizole at a concentration of 1 μM. No significant peaks were detected, and therefore, a "stopped" oscillation condition is reported.

[0149] Figure 21 shows traces from cardiomyocytes treated with nifedipine at a concentration of 1 μM. Only the primary peak 118 is detected, but it is detected at a significantly higher frequency than in the control group.

[0150] Figure 22 shows a compilation and comparison of different readouts for compounds across a normalized concentration range, where the compounds are grouped into three groups—high toxicity, moderate toxicity, and low toxicity—according to known cardiotoxicity. Readouts include the presence of secondary peaks, peak prolongation, irregularity of the interval / amplitude of primary peaks, increased primary peak frequency compared to the control group, termination condition (no effective peak), decreased primary peak frequency compared to the control group, and changes in primary peak amplitude within the oscillation pattern. The graph demonstrates that the appearance of secondary peaks and peak prolongation at concentrations comparable to Cmax (maximum clinical concentration in the blood) is a clear indicator (or correlates) of strong cardiotoxicity, while other readout modifications do not necessarily indicate cardiotoxic effects. (Example 7) Compound testing using neurons

[0151] This example illustrates results obtained by testing neurotoxic compounds in vitro using neurons to assess their effects on various parameters measured by analytical software (see Figures 23-30).

[0152] To accelerate the development of more effective and safer drugs, there is a growing need for more complex, biologically relevant, predictive cell-based assays for drug discovery and toxicology screening. Human iPSC-derived neural 3D co-cultures (StemoniX® microBrain® 3D platform) have been developed as a high-capacity screening platform that more closely resembles the composition of innate human cortical brain tissue. Neural spheroid 3D co-cultures are physiologically relevant co-cultures of functionally active cortical glutamatergic and GABAergic neurons derived from iPSCs that co-differentiate and mature together with astrocytes from the same donor. 3D neural spheroids contain synapse-enriched neural networks that form highly functional neural circuits and exhibit spontaneous, synchronized, and easily detectable calcium oscillations.

[0153] A novel method for the analysis of complex calcium oscillations is disclosed herein. This method enables the detection of oscillation peaks and multi-parameter characterization. Multi-parameter characterization may include oscillation rate (i.e., frequency of the primary peak), width and amplitude of the primary peak, descriptor of the secondary peak, waveform irregularity, and several other important readings.

[0154] Neurons in microBrain® 3D spheroids generate spontaneous, synchronized calcium oscillations. High-speed kinematic fluorescence imaging using the FLIPR® Penta® system allows for intracellular Ca66 assay using the FLIPR® Calcium 6 assay kit. 2+ Ca of neuronal spheroids that are monitored by changes in their levels 2+The patterns and frequencies of the oscillations were measured. A set of known neuromodulators, including NMDA, GABA, and AMPA receptor agonists and antagonists, kainic acid, and analgesics and antiepileptic drugs, were tested. Fluorescence from each well across a 384-well plate was simultaneously recorded at a frequency of 2 Hz (0.5-second sampling interval) by high-speed imaging using the FLIPR® Penta® system.

[0155] The advanced analysis methods implemented in the ScreenWorks® Peak Pro® 2 software module are Ca 2+ This provides a multi-parameter characterization of flux vibration patterns. This phenotypic assay generates readouts of vibration frequency, amplitude, peak width, peak rise and decay times, and peak amplitude / interval irregularity. The effects of neuronal activity moduliators were evaluated by measuring changes in several parameters.

[0156] A set of more than 20 compounds, including several known modulators of neuronal activity, was assayed at different time points and concentrations, and EC50 values ​​were calculated in relation to the compound effects. Time points included 0, 15, 30, 60, 90, and 120 minutes, and 24 hours. Changes were observed as suppression or activation of peak frequencies, or other measurements, corresponding to the expected effects of the corresponding neuromodulators.

[0157] Figures 23-30 show representative traces of calcium oscillations in control and compound-treated neuronal spheroids, recorded over 10 minutes, starting after 30 minutes of treatment with the indicated compounds and concentrations.

[0158] The control group trace (DMSO-treated spheroids) is shown in Figure 23. Only the primary peak 118 (marked with a circle) is detectable. The primary peaks have a somewhat uniform amplitude and spacing from one another.

[0159] Figure 24 shows traces from spheroids treated with MK-801 at a concentration of 3 μM. The primary peak 118 has a variable amplitude (compared to Figure 23). Only one secondary peak 150 is detected.

[0160] Figure 25 shows traces from spheroids treated with GABA at a concentration of 10 μM. Only the primary peak 118 is detected, but it is detected at a significantly lower frequency than in the control group.

[0161] Figure 26 shows traces from spheroids treated with baclofen at a concentration of 10 μM. Only the primary peak 118 is detected, but it is detected at significantly lower frequencies than the control group, with irregular intervals and amplitudes.

[0162] Figure 27 shows traces from spheroids treated with 4-aminopyridine at a concentration of 30 μM. Primary peak 118 and secondary peak 150 are detected. The frequency of primary peak 118 is increased compared to the control group.

[0163] Figure 28 shows traces from spheroids treated with valinomycin at a concentration of (0.3 μM). Primary peak 118 and secondary peak 150 are detected. The frequency of primary peak 118 is much lower than that of the control group, and its amplitude and interval are much more variable.

[0164] Figure 29 shows traces from spheroids treated with kainic acid at a concentration of 1 μM. A primary peak 118 and numerous secondary peaks 150 are detected. The frequency of primary peak 118 is increased compared to the control group.

[0165] Figure 30 shows traces from spheroids treated with tamoxifen at a concentration of 30 μM. Only the primary peak 118 is detected, but it is detected with variable amplitude at a significantly lower frequency than in the control group.

[0166] Assays can be used to test the effects of compounds and screen for neurotoxic chemicals. The appearance of secondary peaks and low-amplitude primary peaks may indicate compounds with neurotoxic activity. Effects on peak percentage and amplitude alone may not adequately predict neurotoxicity. (Example 8) Selected Example

[0167] This embodiment describes selected embodiments of the present disclosure as a series of indexed paragraphs.

[0168] Paragraph 1. A method of analysis comprising: (i) detecting fluorescence representing an oscillatory ion flux associated with one or more biological cells in order to generate a series of data points describing an oscillatory pattern; (ii) calculating a series of gradients relating to the oscillatory pattern; and (iii) using the series of gradients to identify peaks in the oscillatory pattern.

[0169] Paragraph 2. The method according to Paragraph 1, wherein the calculation step uses a sliding window and then defines a subset of data points for which a series of gradients are calculated.

[0170] Paragraph 3. The method according to Paragraph 2, further comprising the step of selecting the size of the sliding window from a set of allowed sizes, wherein the size of the sliding window corresponds to the number of data points from a set of data points to be contained by the sliding window.

[0171] Paragraph 4. The size of the sliding window is automatically assigned by the processor based on the noise level and / or sampling interval of the set of data points in the vibration pattern, and optionally, the processor also calculates a series of gradients and identifies peaks, as described in Paragraph 3.

[0172] Paragraph 5. The step of selecting the size of the sliding window is performed by the user and communicated to the processor, similarly calculating a series of gradients and identifying peaks, as described in Paragraph 3 or 4.

[0173] Paragraph 6. The set of data points is not filtered to reduce noise prior to the step of calculating a series of gradients, as described in any of the methods in Paragraphs 1-5.

[0174] Paragraph 7. The method according to any of paragraphs 1-6, wherein the peak comprises a series of primary peaks, and the oscillation pattern comprises a series of events, each of which comprises only one of the primary peaks, and the method further comprises the step of determining at least one aspect of the secondary peaks of the identified peak, each secondary peak following one of the primary peaks in the event.

[0175] Paragraph 8. The method according to Paragraph 7, wherein at least one aspect of the secondary peaks relates to the number, frequency, or period of secondary peaks in the vibration pattern.

[0176] Paragraph 9. The method according to paragraph 7 or 8, wherein the vibration pattern crosses a predetermined trigger level twice for each event, and the trigger level is set relative to the baseline of the vibration pattern.

[0177] Paragraph 10. The method according to Paragraph 9, wherein the trigger level is set at a certain percentage of the amplitude range over which a series of data points are affected.

[0178] Paragraph 11. The trigger level is the trigger level selected by the user, as described in Paragraph 9.

[0179] Paragraph 12. The method according to any of paragraphs 9-11, wherein the baseline and trigger levels are adjustable by the user via a graphical user interface.

[0180] Paragraph 13. The method according to any of paragraphs 1-12, wherein the step of identifying a peak includes the step of searching for a positive-to-negative or negative-to-positive transition in the gradient within a set of gradients.

[0181] Paragraph 14. The method of Paragraph 13, wherein the step of identifying peaks includes the step of filtering peaks associated with transitions in order to obtain a set of peaks that are considered valid.

[0182] Paragraph 15. The method according to Paragraph 14, wherein the step of filtering peaks includes the step of filtering peaks based on one or more predetermined amplitude and / or duration criteria.

[0183] Paragraph 16. The method according to paragraph 14 or 15, further comprising the step of determining the values ​​of peak-related parameters for a set of peaks that are considered valid.

[0184] Paragraph 17. The method according to any one of paragraphs 14-16, wherein the step of identifying a peak includes: determining the amplitude and / or duration of at least one peak associated with each of a plurality of transitions found by search; comparing the at least one amplitude and / or duration to at least one threshold; and if the comparison step does not satisfy one or more predetermined criteria regarding the at least one amplitude and / or duration, rejecting the peak associated with the transition as invalid.

[0185] Paragraph 18. The method according to Paragraph 17, wherein the step of comparing at least one amplitude and / or duration includes the step of comparing the amplitude with respect to a peak measured relative to a baseline with respect to an oscillation pattern.

[0186] Paragraph 19. The method according to paragraph 17 or 18, wherein the step of comparing at least one amplitude and / or duration includes the step of comparing the local amplitude with respect to a peak measured with respect to a local trough adjacent to the peak.

[0187] Paragraph 20. The method according to any of paragraphs 17-19, wherein the step of comparing at least one amplitude and / or duration includes the step of comparing the durations of peaks measured for at least one local trough.

[0188] Paragraph 21. The method according to any of paragraphs 17-20, wherein each of at least one thresholds is adjustable by the user via a graphical user interface, and / or at least one threshold is set automatically by the processor.

[0189] Paragraph 22. The method according to any one of paragraphs 1-21, wherein the vibration pattern comprises a series of events, each event in the series of events crossing a predetermined trigger level twice, and the method further comprises the step of filtering the series of events to exclude each event having less than a predetermined duration, if present, and each peak in the excluded events is considered invalid.

[0190] Paragraph 23. The method according to any of paragraphs 1-22, further comprising the step of labeling one or more biological cells with a calcium indicator, wherein fluorescence is emitted by the calcium indicator.

[0191] Paragraph 24. The method according to any of paragraphs 1-23, wherein one or more biological cells include one or more cardiomyocytes or neurons.

[0192] Paragraph 25. The method according to paragraph 24, wherein one or more biological cells are primarily cardiomyocytes or neurons.

[0193] Paragraph 26. The method according to paragraph 24 or 25, wherein one or more biological cells include one or more cardiomyocytes or neurons differentiated in vitro from at least one stem cell.

[0194] Paragraph 27. The method according to any of paragraphs 1-26, wherein one or more biological cells are contained in a container selected from petri dishes, flasks, and multiwell microplates.

[0195] Paragraph 28. A series of data points representing a sampling rate greater than 1 Hz, as described in any of paragraphs 1-27.

[0196] Paragraph 29. The method according to any one of paragraphs 1-28, wherein the vibration pattern comprises a series of events, each containing a single primary peak, and the vibration pattern comprises one or more secondary peaks contained in each event of the series of events, and the method further comprises the step of determining at least one value for one or more parameters associated with one or more secondary peaks.

[0197] Paragraph 30. The method according to Paragraph 29, wherein at least one value corresponds to the number, frequency, or period of secondary peaks.

[0198] Paragraph 31. The method according to any of paragraphs 1-30, wherein the steps of detecting fluorescence, calculating a series of gradients, and identifying peaks are performed for each distinct set of multiple distinct sets of one or more biological cells, each distinct set being exposed to a different compound or the same compound at different concentrations.

[0199] Paragraph 32. The method according to paragraph 31, further comprising the step of determining the effect of each different compound or concentration on one or more parameters, if present, that are associated with at least a subset of the peaks, respectively.

[0200] Paragraph 33. The method according to paragraph 32, wherein one or more parameters correspond to the number, frequency, or period of secondary peaks in the vibration pattern.

[0201] Paragraph 34. The method according to paragraph 32 or 33, further comprising the step of predicting the degree of cardiotoxicity or neurotoxicity of each compound or concentration based on its effect on one or more parameters, if present.

[0202] Paragraph 35. The method according to any of paragraphs 1-34, further comprising the step of automatically establishing a baseline for the vibration pattern.

[0203] Paragraph 36. The method according to Paragraph 35, wherein the step of establishing a baseline includes the steps of creating a reference line above the vibration pattern and a threshold line extending below the vibration pattern, finding the maximum value relative to the reference line, wherein the maximum value is located below the threshold line, and performing a linear regression using at least a subset of the maximum values.

[0204] Paragraph 37. The method according to paragraph 36, further comprising the step of calculating the noise level of the vibration pattern using at least a subset of the maximum values.

[0205] Paragraph 38. The step of calculating a series of gradients is to use a sliding window and then define a subset of a series of data points for which a series of gradients are calculated, and the size of the sliding window is selected based on the noise level, as described in any of paragraphs 37.

[0206] Paragraph 39. A method of analysis comprising: (i) detecting fluorescence representing an oscillatory ion flux associated with one or more biological cells in order to generate a series of data points describing an oscillatory pattern; (ii) identifying primary and secondary peaks in the oscillatory pattern, if present; and (iii) determining the aspects of the secondary peaks.

[0207] Paragraph 40. The method according to paragraph 39, wherein the step of determining the aspects of the secondary peaks includes the step of determining the number, frequency, or period of the secondary peaks.

[0208] Paragraph 41. The method according to paragraph 39 or 40, further comprising the steps of exposing one or more biological cells to a compound and determining the effect of the compound on the aspect of a secondary peak.

[0209] Paragraph 42. The method according to any of paragraphs 39-41, wherein the oscillation pattern comprises a series of events, each containing only one of the primary peaks, and each secondary peak occurs after the primary peak in the event.

[0210] Paragraph 43. A secondary peak is a secondary peak generated by cardiomyocytes or neurons, as described in any of paragraphs 39–42.

[0211] Paragraph 44. The method according to any of paragraphs 39-43, further comprising the step of labeling one or more biological cells with a calcium indicator, wherein fluorescence is detected from the calcium indicator.

[0212] Paragraph 45. The method according to any of paragraphs 39-44, further comprising the step of determining the interval regularity / irregularity of the primary peaks.

[0213] Paragraph 46. The method according to Paragraph 45, wherein the step of determining interval regularity / irregularity includes the step of comparing the standard deviation of the intervals of the primary peaks with the mean interval of the primary peaks.

[0214] Paragraph 47. The method according to any of paragraphs 39-46, further comprising the step of determining the amplitude regularity / irregularity of the primary peak.

[0215] Paragraph 48. The method according to Paragraph 47, wherein the step of determining amplitude regularity / irregularity includes the step of comparing the standard deviation of the amplitude of the primary peak with the mean amplitude of the primary peak.

[0216] Paragraph 49. The method according to any of paragraphs 39-48, further comprising the step of comparing the amplitude of each primary peak to a predetermined threshold in order to enumerate any smaller peaks of the primary peak, if any exist.

[0217] Paragraph 50. The method according to any of paragraphs 39–49, wherein the detection step, the identification step, and the determination step are performed for each distinct set of multiple distinct sets of one or more biological cells, and each distinct set is exposed to a different compound or the same compound at different concentrations.

[0218] Paragraph 51. The method according to paragraph 50, further comprising the step of determining the effect of each different compound or concentration on the side of a secondary peak, if present.

[0219] Paragraph 52. The method according to paragraph 50 or 51, further comprising the step of determining the effect of each different compound or concentration on the regularity / irregularity of the spacing of the primary peaks, if present.

[0220] Paragraph 53. The method according to any of paragraphs 50-52, further comprising the step of determining the effect of each different compound or concentration on the amplitude regularity / irregularity of the primary peak, if present.

[0221] Paragraph 54. The method according to any of paragraphs 50-53, further comprising the step of predicting the degree of cardiotoxicity or neurotoxicity of each compound or concentration based on the laterality of secondary peaks and the regularity / irregularity of the spacing of primary peaks.

[0222] Paragraph 55. The method according to any of paragraphs 50-54, further comprising the step of predicting the degree of cardiotoxicity or neurotoxicity of each compound or concentration based on the laterality of the secondary peaks and the amplitude regularity / irregularity of the primary peaks.

[0223] Paragraph 56. The method according to any one of paragraphs 50-55, further comprising the step of predicting the degree of cardiotoxicity or neurotoxicity of each compound or concentration based on the number, frequency, or period of primary peaks, which are secondary peaks and smaller peaks having amplitudes below a predetermined threshold.

[0224] Paragraph 57. A system comprising (i) an optical sensor configured to detect fluorescence representing an oscillatory ion flux associated with one or more biological cells in order to generate a set of data points describing an oscillatory pattern; and (ii) a processor configured to (1) optionally use a sliding window to calculate a set of gradients with respect to an oscillatory pattern, and therefrom define a subset of a set of data points from which a set of gradients is calculated; and (2) use the set of gradients to identify peaks in an oscillatory pattern.

[0225] Paragraph 58. The system described in paragraph 57, configured to carry out any combination of the steps of paragraphs 1-56.

[0226] Paragraph 59. A system comprising (i) an optical sensor configured to detect fluorescence representing an oscillatory ion flux associated with one or more biological cells in order to generate a series of data points describing an oscillatory pattern, and (ii) a processor configured to (1) identify primary and secondary peaks in the oscillatory pattern, if present, and (2) determine the aspects of the secondary peaks.

[0227] Paragraph 60. The system described in paragraph 59, configured to carry out any combination of the steps of paragraphs 1-56.

[0228] As used in this disclosure, the term “exemplary” means “exemplifying” or “serving as an example.” Similarly, the term “exemplify” means “to illustrate by providing an example.” Neither term implies desirability or superiority.

[0229] The disclosures described above may encompass several distinctly different inventions, each with its own independent utility. While each of these inventions is disclosed in its preferred form, the specific embodiments disclosed and illustrated herein are not considered limitingly, as numerous variations are possible. The subject matter of the present invention includes all novel and non-obvious combinations and secondary combinations of the various elements, features, functions, and / or properties disclosed herein.

Claims

1. A method of analysis, wherein the method is: To generate a series of data points (120) describing an oscillation pattern (110), the fluorescence (91) representing the oscillation ion flux associated with one or more biological cells (72) is detected (54), Identifying a plurality of primary peaks (118) and a plurality of secondary peaks (150a, 150b) in the vibration pattern (110) (62), The number, frequency, or period of the aforementioned multiple secondary peaks (150a, 150b) is determined. Includes, The vibration pattern (110) includes a series of events (138), and the vibration pattern (110) crosses a predetermined trigger level (144) twice for each event (138), and the predetermined trigger level (144) is set relative to the baseline (116) of the vibration pattern (110). Each event includes a single primary peak (118), and each of the plurality of secondary peaks (150a, 150b) is included in one of the series of events (138), and each secondary peak is any peak following the primary peak within one event. The detection, identification, and determination are performed on each distinct set of one or more distinct sets of biological cells (72), each distinct set being exposed to different concentrations of different compounds or the same compound. The aforementioned method, Determining the effect of each different compound or each different concentration on the number, frequency, or period of the plurality of secondary peaks (150a, 150b), To predict the degree of toxicity of each compound or each concentration based on the number, frequency, or period of the plurality of secondary peaks, the regularity or irregularity of the intervals or amplitudes of the plurality of primary peaks, or the number, frequency, or period of the plurality of primary peaks that are smaller peaks having amplitudes below a predetermined threshold. Methods that further include the above.

2. The method further comprises determining the regularity or irregularity of the spacing between the plurality of primary peaks (118), and / or the regularity or irregularity of the amplitudes of the plurality of primary peaks (118), or The method according to claim 1, further comprising comparing the amplitude (194) of each primary peak (118) with a predetermined threshold in order to enumerate smaller peaks of the plurality of primary peaks.

3. The method according to claim 1 or claim 2, further comprising labeling one or more biological cells (72) (52) using a calcium indicator, wherein the fluorescence (91) is emitted by the calcium indicator.

4. The method according to claim 1, wherein the degree of toxicity includes the degree of cardiotoxicity or neurotoxicity.

5. A system (70), wherein the system (70) is An optical sensor (88) is configured to detect fluorescence (91) representing the vibrational ion flux associated with one or more biological cells (72) in order to generate a series of data points (120) that describe the vibration pattern (110), A processor (96) is configured to (1) identify a plurality of primary peaks (118) and a plurality of secondary peaks (150a, 150b) in the vibration pattern (110), and (2) determine the number, frequency, or period of the plurality of secondary peaks (150a, 150b) for a vibration pattern (110) that includes a series of events (138). Equipped with, The vibration pattern (110) crosses a predetermined trigger level (144) twice for each event (138), and the predetermined trigger level (144) is set relative to the baseline (116) of the vibration pattern (110). Each event includes a single primary peak (118), and each of the plurality of secondary peaks (150a, 150b) is included in one of the series of events (138), and each secondary peak is any peak following the primary peak within one event. The detection, identification, and determination are performed on each distinct set of one or more distinct sets of biological cells (72), each distinct set being exposed to different concentrations of different compounds or the same compound. The aforementioned processor (96) Determining the effect of each different compound or each different concentration on the number, frequency, or period of the plurality of secondary peaks (150a, 150b), To predict the degree of toxicity of each compound or each concentration based on the number, frequency, or period of the plurality of secondary peaks, the regularity or irregularity of the intervals or amplitudes of the plurality of primary peaks, or the number, frequency, or period of the plurality of primary peaks that are smaller peaks having amplitudes below a predetermined threshold. The system (70) is further configured to perform the following actions.

Images