-
Empirical quantification of predictive uncertainty due to model discrepancy by training with an ensemble of experimental designs: an application to ion channel kinetics
Authors:
Joseph G. Shuttleworth,
Chon Lok Lei,
Dominic G. Whittaker,
Monique J. Windley,
Adam P. Hill,
Simon P. Preston,
Gary R. Mirams
Abstract:
When mathematical biology models are used to make quantitative predictions for clinical or industrial use, it is important that these predictions come with a reliable estimate of their accuracy (uncertainty quantification). Because models of complex biological systems are always large simplifications, model discrepancy arises - where a mathematical model fails to recapitulate the true data generat…
▽ More
When mathematical biology models are used to make quantitative predictions for clinical or industrial use, it is important that these predictions come with a reliable estimate of their accuracy (uncertainty quantification). Because models of complex biological systems are always large simplifications, model discrepancy arises - where a mathematical model fails to recapitulate the true data generating process. This presents a particular challenge for making accurate predictions, and especially for making accurate estimates of uncertainty in these predictions. Experimentalists and modellers must choose which experimental procedures (protocols) are used to produce data to train their models. We propose to characterise uncertainty owing to model discrepancy with an ensemble of parameter sets, each of which results from training to data from a different protocol. The variability in predictions from this ensemble provides an empirical estimate of predictive uncertainty owing to model discrepancy, even for unseen protocols. We use the example of electrophysiology experiments, which are used to investigate the kinetics of the hERG potassium ion channel. Here, 'information-rich' protocols allow mathematical models to be trained using numerous short experiments performed on the same cell. Typically, assuming independent observational errors and training a model to an individual experiment results in parameter estimates with very little dependence on observational noise. Moreover, parameter sets arising from the same model applied to different experiments often conflict - indicative of model discrepancy. Our methods will help select more suitable mathematical models of hERG for future studies, and will be widely applicable to a range of biological modelling problems.
△ Less
Submitted 19 February, 2024; v1 submitted 6 February, 2023;
originally announced February 2023.
-
Structural modeling of 2019-novel coronavirus (nCoV) spike protein reveals a proteolytically-sensitive activation loop as a distinguishing feature compared to SARS-CoV and related SARS-like coronaviruses
Authors:
Javier A. Jaimes,
Nicole M. Andre,
Jean K. Millet,
Gary R. Whittaker
Abstract:
The 2019 novel coronavirus (2019-nCoV) is currently causing a widespread outbreak centered on Hubei province, China and is a major public health concern. Taxonomically 2019-nCoV is closely related to SARS-CoV and SARS-related bat coronaviruses, and it appears to share a common receptor with SARS-CoV (ACE-2). Here, we perform structural modeling of the 2019-nCoV spike glycoprotein. Our data provide…
▽ More
The 2019 novel coronavirus (2019-nCoV) is currently causing a widespread outbreak centered on Hubei province, China and is a major public health concern. Taxonomically 2019-nCoV is closely related to SARS-CoV and SARS-related bat coronaviruses, and it appears to share a common receptor with SARS-CoV (ACE-2). Here, we perform structural modeling of the 2019-nCoV spike glycoprotein. Our data provide support for the similar receptor utilization between 2019-nCoV and SARS-CoV, despite a relatively low amino acid similarity in the receptor binding module. Compared to SARS-CoV, we identify an extended structural loop containing basic amino acids at the interface of the receptor binding (S1) and fusion (S2) domains, which we predict to be proteolytically-sensitive. We suggest this loop confers fusion activation and entry properties more in line with MERS-CoV and other coronaviruses, and that the presence of this structural loop in 2019-nCoV may affect virus stability and transmission.
△ Less
Submitted 13 February, 2020;
originally announced February 2020.
-
Considering discrepancy when calibrating a mechanistic electrophysiology model
Authors:
Chon Lok Lei,
Sanmitra Ghosh,
Dominic G. Whittaker,
Yasser Aboelkassem,
Kylie A. Beattie,
Chris D. Cantwell,
Tammo Delhaas,
Charles Houston,
Gustavo Montes Novaes,
Alexander V. Panfilov,
Pras Pathmanathan,
Marina Riabiz,
Rodrigo Weber dos Santos,
John Walmsley,
Keith Worden,
Gary R. Mirams,
Richard D. Wilkinson
Abstract:
Uncertainty quantification (UQ) is a vital step in using mathematical models and simulations to take decisions. The field of cardiac simulation has begun to explore and adopt UQ methods to characterise uncertainty in model inputs and how that propagates through to outputs or predictions. In this perspective piece we draw attention to an important and under-addressed source of uncertainty in our pr…
▽ More
Uncertainty quantification (UQ) is a vital step in using mathematical models and simulations to take decisions. The field of cardiac simulation has begun to explore and adopt UQ methods to characterise uncertainty in model inputs and how that propagates through to outputs or predictions. In this perspective piece we draw attention to an important and under-addressed source of uncertainty in our predictions -- that of uncertainty in the model structure or the equations themselves. The difference between imperfect models and reality is termed model discrepancy, and we are often uncertain as to the size and consequences of this discrepancy. Here we provide two examples of the consequences of discrepancy when calibrating models at the ion channel and action potential scales. Furthermore, we attempt to account for this discrepancy when calibrating and validating an ion channel model using different methods, based on modelling the discrepancy using Gaussian processes (GPs) and autoregressive-moving-average (ARMA) models, then highlight the advantages and shortcomings of each approach. Finally, suggestions and lines of enquiry for future work are provided.
△ Less
Submitted 23 April, 2020; v1 submitted 13 January, 2020;
originally announced January 2020.
-
Studying the photometric and spectroscopic variability of the magnetic hot supergiant $ζ$ Orionis Aa
Authors:
B. Buysschaert,
C. Neiner,
N. D. Richardson,
T. Ramiaramanantsoa,
A. David-Uraz,
H. Pablo,
M. E. Oksala,
A. F. J. Moffat,
R. E. Mennickent,
S. Legeza,
C. Aerts,
R. Kuschnig,
G. N. Whittaker,
A. Popowicz,
G. Handler,
G. A. Wade,
W. W. Weiss
Abstract:
Massive stars play a significant role in the chemical and dynamical evolution of galaxies. However, much of their variability, particularly during their evolved supergiant stage, is poorly understood. To understand the variability of evolved massive stars in more detail, we present a study of the O9.2Ib supergiant $ζ$ Ori Aa, the only currently confirmed supergiant to host a magnetic field. We hav…
▽ More
Massive stars play a significant role in the chemical and dynamical evolution of galaxies. However, much of their variability, particularly during their evolved supergiant stage, is poorly understood. To understand the variability of evolved massive stars in more detail, we present a study of the O9.2Ib supergiant $ζ$ Ori Aa, the only currently confirmed supergiant to host a magnetic field. We have obtained two-color space-based BRIght Target Explorer photometry (BRITE) for $ζ$ Ori Aa during two observing campaigns, as well as simultaneous ground-based, high-resolution optical CHIRON spectroscopy. We perform a detailed frequency analysis to detect and characterize the star's periodic variability. We detect two significant, independent frequencies, their higher harmonics, and combination frequencies: the stellar rotation period $P_{\mathrm{rot}} = 6.82\pm0.18$ d, most likely related to the presence of the stable magnetic poles, and a variation with a period of $10.0\pm0.3$ d attributed to circumstellar environment, also detected in the H$α$ and several He I lines, yet absent in the purely photospheric lines. We confirm the variability with $P_{\mathrm{rot}}$/4, likely caused by surface inhomogeneities, being the possible photospheric drivers of the discrete absorption components. No stellar pulsations were detected in the data. The level of circumstellar activity clearly differs between the two BRITE observing campaigns. We demonstrate that $ζ$ Ori Aa is a highly variable star with both periodic and non-periodic variations, as well as episodic events. The rotation period we determined agrees well with the spectropolarimetric value from the literature. The changing activity level observed with BRITE could explain why the rotational modulation of the magnetic measurements was not clearly detected at all epochs.
△ Less
Submitted 3 April, 2017;
originally announced April 2017.
-
STEREO observations of HD 90386 (RX Sex): a Delta Scuti or a hybrid star?
Authors:
Dogus Ozuyar,
Ian R. Stevens,
Gemma Whittaker,
Vinothini Sangaralingam
Abstract:
HD 90386 is a rarely studied bright A2V type Delta Scuti star (V = 6.66 mag). It displays short-term light curve variations which are originated due to either a beating phenomenon or a non-periodic variation. In this paper, we presented high-precision photometric data of HD 90386 taken by the STEREO satellite between 2007 and 2011 to shed light on its internal structure and evolution stage. From t…
▽ More
HD 90386 is a rarely studied bright A2V type Delta Scuti star (V = 6.66 mag). It displays short-term light curve variations which are originated due to either a beating phenomenon or a non-periodic variation. In this paper, we presented high-precision photometric data of HD 90386 taken by the STEREO satellite between 2007 and 2011 to shed light on its internal structure and evolution stage. From the frequency analysis of the four-year data, we detected that HD 90386 had at least six different frequencies between 1 and 15 c~d$^{-1}$. The most dominant frequencies were found at around 10.25258 c~d$^{-1}$ (A $\sim$ 1.92 mmag) and 12.40076 c~d$^{-1}$ (A $\sim$ 0.61 mmag). Based on the ratio between these frequencies, the star was considered as an overtone pulsator. The variation in pulsation period over 35 years was calculated to be dP/Pdt = 5.39(2) x 10$^{-3}$ yr$^{-1}$. Other variabilities at around 1.0 c~d$^{-1}$ in the amplitude spectrum of HD90386 were also discussed. In order to explain these variabilities, possible rotational effects and Gamma Dor type variations were focused. Consequently, depending on the rotation velocity of HD 90386, we speculated that these changes might be related to Gamma Dor type high-order g-modes shifted to the higher frequencies and that HD 90386 might be a hybrid star.
△ Less
Submitted 31 January, 2017;
originally announced January 2017.
-
Combining BRITE and ground-based photometry for the Beta Cephei star Nu Eridani: impact on photometric pulsation mode identification and detection of several g modes
Authors:
G. Handler,
M. Rybicka,
A. Popowicz,
A. Pigulski,
R. Kuschnig,
E. Zoclonska,
A. F. J. Moffat,
W. W. Weiss,
C. C. Grant,
H. Pablo,
G. N. Whittaker,
S. M. Rucinski,
T. Ramiaramanantsoa,
K. Zwintz,
G. A. Wade
Abstract:
We report a simultaneous ground and space-based photometric study of the Beta Cephei star Nu Eridani. Half a year of observations have been obtained by four of the five satellites constituting BRITE-Constellation, supplemented with ground-based photoelectric photometry. We show that carefully combining the two data sets virtually eliminates the aliasing problem that often hampers time-series analy…
▽ More
We report a simultaneous ground and space-based photometric study of the Beta Cephei star Nu Eridani. Half a year of observations have been obtained by four of the five satellites constituting BRITE-Constellation, supplemented with ground-based photoelectric photometry. We show that carefully combining the two data sets virtually eliminates the aliasing problem that often hampers time-series analyses. We detect 40 periodic signals intrinsic to the star in the light curves. Despite a lower detection limit we do not recover all the pressure and mixed modes previously reported in the literature, but we newly detect six additional gravity modes. This behaviour is a consequence of temporal changes in the pulsation amplitudes that we also detected for some of the p modes. We point out that the dependence of theoretically predicted pulsation amplitude on wavelength is steeper in visual passbands than those observationally measured, to the extent that the three dominant pulsation modes of Nu Eridani would be incorrectly identified using data in optical filters only. We discuss possible reasons for this discrepancy.
△ Less
Submitted 29 September, 2016;
originally announced September 2016.
-
The BRITE Constellation nanosatellite mission: Testing, commissioning and operations
Authors:
H. Pablo,
G. N. Whittaker,
A. Popowicz,
S. M. Mochnacki,
R. Kuschnig,
C. C. Grant,
A. F. J. Moffat,
S. M. Rucinski,
J. M. Matthews,
A. Schwarzenberg-Czerny,
G. Handler,
W. W. Weiss,
D. Baade,
G. A. Wade,
E. Zoclonska,
T. Ramiaramanantsoa,
M. Unterberger,
K. Zwintz,
A. Pigulski,
J. Rowe,
O. Koudelka,
P. Orleanski,
A. Pamyatnykh,
C. Neiner,
R. Wawrzaszek
, et al. (5 additional authors not shown)
Abstract:
BRITE (BRIght Target Explorer) Constellation, the first nanosatellite mission applied to astrophysical research, is a collaboration among Austria, Canada and Poland. The fleet of satellites (6 launched, 5 functioning) performs precise optical photometry of the brightest stars in the night sky. A pioneering mission like BRITE - with optics and instruments restricted to small volume, mass and power…
▽ More
BRITE (BRIght Target Explorer) Constellation, the first nanosatellite mission applied to astrophysical research, is a collaboration among Austria, Canada and Poland. The fleet of satellites (6 launched, 5 functioning) performs precise optical photometry of the brightest stars in the night sky. A pioneering mission like BRITE - with optics and instruments restricted to small volume, mass and power in several nanosatellites, whose measurements must be coordinated in orbit - poses many unique challenges. We discuss the technical issues, including problems encountered during on-orbit commissioning (especially higher-than expected sensitivity of the CCDs to particle radiation). We describe in detail how the BRITE team has mitigated these problems, and provide a complete overview of mission operations. This paper serves as a template for how to effectively plan, build and operate future low-cost niche-driven space astronomy missions.
△ Less
Submitted 31 July, 2016;
originally announced August 2016.
-
Massive pulsating stars observed by BRITE-Constellation. I. The triple system Beta Centauri (Agena)
Authors:
A. Pigulski,
H. Cugier,
A. Popowicz,
R. Kuschnig,
A. F. J. Moffat,
S. M. Rucinski,
A. Schwarzenberg-Czerny,
W. W. Weiss,
G. Handler,
G. A. Wade,
O. Koudelka,
J. M. Matthews,
St. Mochnacki,
P. Orleański,
H. Pablo,
T. Ramiaramanantsoa,
G. Whittaker,
E. Zocłońska,
K. Zwintz
Abstract:
This paper aims to precisely determine the masses and detect pulsation modes in the two massive components of Beta Cen with BRITE-Constellation photometry. In addition, seismic models for the components are considered and the effects of fast rotation are discussed. This is done to test the limitations of seismic modeling for this very difficult case. A simultaneous fit of visual and spectroscopic…
▽ More
This paper aims to precisely determine the masses and detect pulsation modes in the two massive components of Beta Cen with BRITE-Constellation photometry. In addition, seismic models for the components are considered and the effects of fast rotation are discussed. This is done to test the limitations of seismic modeling for this very difficult case. A simultaneous fit of visual and spectroscopic orbits is used to self-consistently derive the orbital parameters, and subsequently the masses, of the components. The derived masses are equal to 12.02 +/- 0.13 and 10.58 +/- 0.18 M_Sun. The parameters of the wider, A - B system, presently approaching periastron passage, are constrained. Analysis of the combined blue- and red-filter BRITE-Constellation photometric data of the system revealed the presence of 19 periodic terms, of which eight are likely g modes, nine are p modes, and the remaining two are combination terms. It cannot be excluded that one or two low-frequency terms are rotational frequencies. It is possible that both components of Beta Cen are Beta Cep/SPB hybrids. An attempt to use the apparent changes of frequency to distinguish which modes originate in which component did not succeed, but there is potential for using this method when more BRITE data become available. Agena seems to be one of very few rapidly rotating massive objects with rich p- and g-mode spectra, and precisely known masses. It can therefore be used to gain a better understanding of the excitation of pulsations in relatively rapidly rotating stars and their seismic modeling. Finally, this case illustrates the potential of BRITE-Constellation data for the detection of rich-frequency spectra of small-amplitude modes in massive pulsating stars.
△ Less
Submitted 8 February, 2016;
originally announced February 2016.
-
Short-term variability and mass loss in Be stars I. BRITE satellite photometry of $η$ and $μ$ Centauri
Authors:
D. Baade,
Th. Rivinius,
A. Pigulski,
A. C. Carciofi,
Ch. Martayan,
A. F. J. Moffat,
G. A. Wade,
W. W. Weiss,
J. Grunhut,
G. Handler,
R. Kuschnig,
A. Mehner,
H. Pablo,
A. Popowicz,
S. Rucinski,
G. Whittaker
Abstract:
Empirical evidence for the involvement of nonradial pulsations (NRP's) in the mass loss from Be stars ranges from (i) a singular case (\object{$μ$ Cen}) of repetitive mass ejections triggered by multi-mode beating to (ii) several photometric reports about enormous numbers of pulsation modes pop** up during outbursts and on to (iii) effective single-mode pulsators.
The BRITE Constellation of na…
▽ More
Empirical evidence for the involvement of nonradial pulsations (NRP's) in the mass loss from Be stars ranges from (i) a singular case (\object{$μ$ Cen}) of repetitive mass ejections triggered by multi-mode beating to (ii) several photometric reports about enormous numbers of pulsation modes pop** up during outbursts and on to (iii) effective single-mode pulsators.
The BRITE Constellation of nanosatellites was used to obtain mmag photometry of the Be stars $η$ and \object{$μ$ Cen}.
In the low-inclination star \object{$μ$ Cen}, light pollution by variable amounts of near-stellar matter prevented any new insights into the variability and other properties of the central star. In the equator-on star \object{$η$ Cen}, BRITE photometry and {\sc Heros} echelle spectroscopy from the 1990s reveal an intricate clockwork of star-disk interactions. The mass transfer is modulated with the frequency difference of two NRP modes and an amplitude three times as large as the amplitude sum of the two NRP modes. This process feeds a high-amplitude circumstellar activity running with the incoherent and slightly lower so-called Štefl frequency. The mass loss-modulation cycles are tightly coupled to variations in the value of the Štefl frequency and in its amplitude, albeit with strongly drifting phase differences.
The observations are well described by the decomposition of the mass loss into a pulsation-related engine in the star and a viscosity-dominated engine in the circumstellar disk. Arguments are developed that large-scale gas-circulation flows occur at the interface. The propagation rates of these eddies manifest themselves as Štefl frequencies. Bursts in power spectra during mass-loss events can be understood as the noise inherent to these gas flows.
△ Less
Submitted 4 February, 2016;
originally announced February 2016.
-
The roAp star $α$ Circini as seen by BRITE-Constellation
Authors:
W. W. Weiss,
H. -E. Fröhlich,
A. Pigulski,
A. Popowicz,
D. Huber,
R. Kuschnig,
A. F. J. Moffat,
J. M. Matthews,
H. Saio,
A. Schwarzenberg-Czerny,
C. C. Grant,
O. Koudelka,
T. Lüftinger,
S. M. Rucinski,
G. A. Wade,
J. Alves,
M. Guedel,
G. Handler,
St. Mochnacki,
P. Orleanski,
B. Pablo,
A. Pamyatnykh,
T. Ramiaramanantsoa,
J. Rowe,
G. Whittaker
, et al. (3 additional authors not shown)
Abstract:
We report on an analysis of high-precision, multi-colour photometric observations of the rapidly-oscillating Ap (roAp) star $α$ Cir. These observations were obtained with the BRITE-Constellation, which is a coordinated mission of five nanosatellites that collects continuous millimagnitude-precision photometry of dozens of bright stars for up to 180 days at a time in two colours (Johnson B and R).…
▽ More
We report on an analysis of high-precision, multi-colour photometric observations of the rapidly-oscillating Ap (roAp) star $α$ Cir. These observations were obtained with the BRITE-Constellation, which is a coordinated mission of five nanosatellites that collects continuous millimagnitude-precision photometry of dozens of bright stars for up to 180 days at a time in two colours (Johnson B and R). BRITE stands for BRight Target Explorer. The object $α$ Cir is the brightest roAp star and an ideal target for such investigations, facilitating the determination of oscillation frequencies with high resolution. This star is bright enough for complementary interferometry and time-resolved spectroscopy. Four BRITE satellites observed $α$ Cir for 146 d or 33 rotational cycles. Phasing the photometry according to the 4.4790 d rotational period reveals qualitatively different light variations in the two photometric bands. The phased red-band photometry is in good agreement with previously-published WIRE data, showing a light curve symmetric about phase 0.5 with a strong contribution from the first harmonic. The phased blue-band data, in contrast, show an essentially sinusoidal variation. We model both light curves with Bayesian Photometric Imaging, which suggests the presence of two large-scale, photometrically bright (relative to the surrounding photosphere) spots. We also examine the high-frequency pulsation spectrum as encoded in the BRITE photometry. Our analysis establishes the stability of the main pulsation frequency over the last 20 years, confirms the presence of frequency f7, which was not detected (or the mode not excited) prior to 2006, and excludes quadrupolar modes for the main pulsation frequency.
△ Less
Submitted 19 January, 2016;
originally announced January 2016.
-
Feline coronaviruses associated with feline infectious peritonitis have modifications to spike protein activation sites at two discrete positions
Authors:
Beth N. Licitra,
Kelly L. Sams,
Donald W. Lee,
Gary R. Whittaker
Abstract:
Feline infectious peritonitis (FIP) is associated with mutations in the feline coronavirus (FCoV) genome that are thought to convert the subclinical feline enteric coronavirus (FECV) into the lethal feline infectious peritonitis virus (FIPV). A key feature of FIPV, not shared with FECV, is the productive infection of macrophages. Therefore mutations in proteins that govern cell tropism, such as th…
▽ More
Feline infectious peritonitis (FIP) is associated with mutations in the feline coronavirus (FCoV) genome that are thought to convert the subclinical feline enteric coronavirus (FECV) into the lethal feline infectious peritonitis virus (FIPV). A key feature of FIPV, not shared with FECV, is the productive infection of macrophages. Therefore mutations in proteins that govern cell tropism, such as the spike glycoprotein, may play an important role in FIP progression. In a prior study, involving a limited number of samples, we have shown an association of FIP with mutations in the protease cleavage-activation site located between the receptor-binding and fusion domains of the FCoV spike (S1/S2). Here, we extend these studies to investigate a larger sample set and to obtain a more refined analysis of the mutations at this S1/S2 site. Our larger data set more clearly shows that the mutations acquired by FIPV at S1/S2 are also accompanied by additional mutations at a second protease cleavage-activation site located in the fusion domain (S2'), adjacent to the viral fusion peptide. Overall, our data indicate a pattern of mutations across the two protease recognition sites that results in substitutions, and/or altered recognition, of critical basic/polar amino acid residues needed for virus activation in the enteric tract. Typically, FIPVs have substitutions of non-polar, aliphatic or aromatic residues in the protease recognition sites. These changes likely modulate the proteolytic activation of the virus and its ability to productively infect macrophages in vivo.
△ Less
Submitted 12 December, 2014;
originally announced December 2014.
