-
Assessing Reusability of Deep Learning-Based Monotherapy Drug Response Prediction Models Trained with Omics Data
Authors:
Jamie C. Overbeek,
Alexander Partin,
Thomas S. Brettin,
Nicholas Chia,
Oleksandr Narykov,
Priyanka Vasanthakumari,
Andreas Wilke,
Yitan Zhu,
Austin Clyde,
Sara Jones,
Rohan Gnanaolivu,
Yuanhang Liu,
Jun Jiang,
Chen Wang,
Carter Knutson,
Andrew McNaughton,
Neeraj Kumar,
Gayara Demini Fernando,
Souparno Ghosh,
Cesar Sanchez-Villalobos,
Ruibo Zhang,
Ranadip Pal,
M. Ryan Weil,
Rick L. Stevens
Abstract:
Cancer drug response prediction (DRP) models present a promising approach towards precision oncology, tailoring treatments to individual patient profiles. While deep learning (DL) methods have shown great potential in this area, models that can be successfully translated into clinical practice and shed light on the molecular mechanisms underlying treatment response will likely emerge from collabor…
▽ More
Cancer drug response prediction (DRP) models present a promising approach towards precision oncology, tailoring treatments to individual patient profiles. While deep learning (DL) methods have shown great potential in this area, models that can be successfully translated into clinical practice and shed light on the molecular mechanisms underlying treatment response will likely emerge from collaborative research efforts. This highlights the need for reusable and adaptable models that can be improved and tested by the wider scientific community. In this study, we present a scoring system for assessing the reusability of prediction DRP models, and apply it to 17 peer-reviewed DL-based DRP models. As part of the IMPROVE (Innovative Methodologies and New Data for Predictive Oncology Model Evaluation) project, which aims to develop methods for systematic evaluation and comparison DL models across scientific domains, we analyzed these 17 DRP models focusing on three key categories: software environment, code modularity, and data availability and preprocessing. While not the primary focus, we also attempted to reproduce key performance metrics to verify model behavior and adaptability. Our assessment of 17 DRP models reveals both strengths and shortcomings in model reusability. To promote rigorous practices and open-source sharing, we offer recommendations for developing and sharing prediction models. Following these recommendations can address many of the issues identified in this study, improving model reusability without adding significant burdens on researchers. This work offers the first comprehensive assessment of reusability and reproducibility across diverse DRP models, providing insights into current model sharing practices and promoting standards within the DRP and broader AI-enabled scientific research community.
△ Less
Submitted 18 September, 2024;
originally announced September 2024.
-
Deep learning methods for drug response prediction in cancer: predominant and emerging trends
Authors:
Alexander Partin,
Thomas S. Brettin,
Yitan Zhu,
Oleksandr Narykov,
Austin Clyde,
Jamie Overbeek,
Rick L. Stevens
Abstract:
Cancer claims millions of lives yearly worldwide. While many therapies have been made available in recent years, by in large cancer remains unsolved. Exploiting computational predictive models to study and treat cancer holds great promise in improving drug development and personalized design of treatment plans, ultimately suppressing tumors, alleviating suffering, and prolonging lives of patients.…
▽ More
Cancer claims millions of lives yearly worldwide. While many therapies have been made available in recent years, by in large cancer remains unsolved. Exploiting computational predictive models to study and treat cancer holds great promise in improving drug development and personalized design of treatment plans, ultimately suppressing tumors, alleviating suffering, and prolonging lives of patients. A wave of recent papers demonstrates promising results in predicting cancer response to drug treatments while utilizing deep learning methods. These papers investigate diverse data representations, neural network architectures, learning methodologies, and evaluations schemes. However, deciphering promising predominant and emerging trends is difficult due to the variety of explored methods and lack of standardized framework for comparing drug response prediction models. To obtain a comprehensive landscape of deep learning methods, we conducted an extensive search and analysis of deep learning models that predict the response to single drug treatments. A total of 60 deep learning-based models have been curated and summary plots were generated. Based on the analysis, observable patterns and prevalence of methods have been revealed. This review allows to better understand the current state of the field and identify major challenges and promising solution paths.
△ Less
Submitted 17 November, 2022;
originally announced November 2022.
-
Theorie der Electrophorese -- Het Relaxatie-Effect
Authors:
J. Th. G. Overbeek
Abstract:
In this thesis, a theoretical treatment of the relation between electrophoretic velocity and the potential of the double layer of colloidal particles is presented.
Translators' note: The theory of electrophoresis is one of the foundational topics that underpinned the development of colloid and surface science and ranks with the famous Derjaguin-Landau-Verwey-Overbeek (DLVO) theory of colloidal…
▽ More
In this thesis, a theoretical treatment of the relation between electrophoretic velocity and the potential of the double layer of colloidal particles is presented.
Translators' note: The theory of electrophoresis is one of the foundational topics that underpinned the development of colloid and surface science and ranks with the famous Derjaguin-Landau-Verwey-Overbeek (DLVO) theory of colloidal stability. J. Th. G. Overbeek ("Theo" to all who knew him) was the first to develop a complete theoretical analysis of the electrophoretic motion of a charged spherical particle under the influence of an external electric field. This provided the theoretical framework for a widely used experimental method to characterize the state of charge and particle size of small colloidal particles. The solution of this problem required mastery of fluid mechanics, colloidal electrostatics, statistical thermodynamics and transport theory in addition to solid applied mathematics. Theo carried out this study as his doctoral thesis under H.R. Kruyt at Utrecht University. The thesis, in Dutch, was later published as a monograph. Given the important pedagogic value and historical status of this work, we felt that it deserved to enjoy a wide readership.
△ Less
Submitted 11 July, 2019;
originally announced July 2019.
-
The Gaia-ESO Survey: radial distribution of abundances in the Galactic disc from open clusters and young field stars
Authors:
L. Magrini,
S. Randich,
G. Kordopatis,
N. Prantzos,
D. Romano,
A. Chieffi,
M. Limongi,
P. Francois,
E. Pancino,
E. Friel,
A. Bragaglia,
G. Tautvaisiene,
L. Spina,
J. Overbeek,
T. Cantat-Gaudin,
P. Donati,
A. Vallenari,
R. Sordo,
F. M. Jimenez-Esteban,
B. Tang,
A. Drazdauskas,
S. Sousa,
S. Duffau,
P. Jofre,
G. Gilmore
, et al. (22 additional authors not shown)
Abstract:
The spatial distribution of elemental abundances in the disc of our Galaxy gives insights both on its assembly process and subsequent evolution, and on the stellar nucleogenesis of the different elements. Gradients can be traced using several types of objects as, for instance, (young and old) stars, open clusters, HII regions, planetary nebulae. We aim at tracing the radial distributions of abunda…
▽ More
The spatial distribution of elemental abundances in the disc of our Galaxy gives insights both on its assembly process and subsequent evolution, and on the stellar nucleogenesis of the different elements. Gradients can be traced using several types of objects as, for instance, (young and old) stars, open clusters, HII regions, planetary nebulae. We aim at tracing the radial distributions of abundances of elements produced through different nucleosynthetic channels -the alpha-elements O, Mg, Si, Ca and Ti, and the iron-peak elements Fe, Cr, Ni and Sc - by using the Gaia-ESO idr4 results of open clusters and young field stars. From the UVES spectra of member stars, we determine the average composition of clusters with ages >0.1 Gyr. We derive statistical ages and distances of field stars. We trace the abundance gradients using the cluster and field populations and we compare them with a chemo-dynamical Galactic evolutionary model. Results. The adopted chemo-dynamical model, with the new generation of metallicity-dependent stellar yields for massive stars, is able to reproduce the observed spatial distributions of abundance ratios, in particular the abundance ratios of [O/Fe] and [Mg/Fe] in the inner disc (5 kpc<RGC <7 kpc), with their differences, that were usually poorly explained by chemical evolution models. Often, oxygen and magnesium are considered as equivalent in tracing alpha-element abundances and in deducing, e.g., the formation time-scales of different Galactic stellar populations. In addition, often [alpha/Fe] is computed combining several alpha-elements. Our results indicate, as expected, a complex and diverse nucleosynthesis of the various alpha-elements, in particular in the high metallicity regimes, pointing towards a different origin of these elements and highlighting the risk of considering them as a single class with common features.
△ Less
Submitted 2 March, 2017;
originally announced March 2017.
-
The Gaia-ESO Survey: the inner disk, intermediate-age open cluster Trumpler 23
Authors:
J. C. Overbeek,
E. D. Friel,
P. Donati,
R. Smiljanic,
H. R. Jacobson,
D. Hatzidimitriou,
E. V. Held,
L. Magrini,
A. Bragaglia,
S. Randich,
A. Vallenari,
T. Cantat-Gaudin,
G. Tautvaisiene,
F. Jimenez-Esteban,
A. Frasca,
D. Geisler,
S. Villanova,
B. Tang,
C. Munoz,
G. Marconi,
G. Carraro,
I. San Roman,
A. Drazdauskas,
R. Zenoviene,
G. Gilmore
, et al. (11 additional authors not shown)
Abstract:
Context: Trumpler 23 is a moderately populated, intermediate-age open cluster within the solar circle at a Rgc ~6 kpc. It is in a crowded field very close to the Galactic plane and the color-magnitude diagram shows significant field contamination and possible differential reddening; it is a relatively understudied cluster for these reasons, but its location makes it a key object for determining Ga…
▽ More
Context: Trumpler 23 is a moderately populated, intermediate-age open cluster within the solar circle at a Rgc ~6 kpc. It is in a crowded field very close to the Galactic plane and the color-magnitude diagram shows significant field contamination and possible differential reddening; it is a relatively understudied cluster for these reasons, but its location makes it a key object for determining Galactic abundance distributions.
Aims: New data from the Gaia-ESO Survey enable the first ever radial velocity and spectroscopic metallicity measurements for this cluster. We aim to use velocities to isolate cluster members, providing more leverage for determining cluster parameters.
Methods: Gaia-ESO Survey data for 167 potential members have yielded radial velocity measurements, which were used to determine the systemic velocity of the cluster and membership of individual stars. Atmospheric parameters were also used as a check on membership when available. Literature photometry was used to re-determine cluster parameters based on radial velocity member stars only; theoretical isochrones are fit in the V, V-I diagram. Cluster abundance measurements of ten radial-velocity member stars with high-resolution spectroscopy are presented for 24 elements. These abundances have been compared to local disk stars, and where possible placed within the context of literature gradient studies.
Results: We find Trumpler 23 to have an age of 0.80 +/- 0.10 Gyr, significant differential reddening with an estimated mean cluster E(V-I) of 1.02 +0.14/-0.09, and an apparent distance modulus of 14.15 +/- 0.20. We find an average cluster metallicity of [Fe/H] = 0.14 +/- 0.03 dex, a solar [alpha/Fe] abundance, and notably subsolar [s-process/Fe] abundances.
△ Less
Submitted 10 January, 2017; v1 submitted 2 November, 2016;
originally announced November 2016.
-
The Gaia-ESO Survey: Probes of the inner disk abundance gradient
Authors:
H. R. Jacobson,
E. D. Friel,
L. Jilkova,
L. Magrini,
A. Bragaglia,
A. Vallenari,
M. Tosi,
S. Randich,
P. Donati,
T. Cantat-Gaudin,
R. Sordo,
R. Smiljanic,
J. C. Overbeek,
G. Carraro,
G. Tautvaisiene,
I. San Roman,
S. Villanova,
D. Geisler,
C. Munoz,
F. Jimenez-Esteban,
B. Tang,
G. Gilmore,
E. J. Alfaro,
T. Bensby,
E. Flaccomio
, et al. (19 additional authors not shown)
Abstract:
The nature of the metallicity gradient inside the solar circle (R_GC < 8 kpc) is poorly understood, but studies of Cepheids and a small sample of open clusters suggest that it steepens in the inner disk. We investigate the metallicity gradient of the inner disk using a sample of inner disk open clusters that is three times larger than has previously been studied in the literature to better charact…
▽ More
The nature of the metallicity gradient inside the solar circle (R_GC < 8 kpc) is poorly understood, but studies of Cepheids and a small sample of open clusters suggest that it steepens in the inner disk. We investigate the metallicity gradient of the inner disk using a sample of inner disk open clusters that is three times larger than has previously been studied in the literature to better characterize the gradient in this part of the disk. We used the Gaia-ESO Survey (GES) [Fe/H] values and stellar parameters for stars in 12 open clusters in the inner disk from GES-UVES data. Cluster mean [Fe/H] values were determined based on a membership analysis for each cluster. Where necessary, distances and ages to clusters were determined via comparison to theoretical isochrones. The GES open clusters exhibit a radial metallicity gradient of -0.10+-0.02 dex/kpc, consistent with the gradient measured by other literature studies of field red giant stars and open clusters in the range R_GC ~ 6-12 kpc. We also measure a trend of increasing [Fe/H] with increasing cluster age, as has also been found in the literature. We find no evidence for a steepening of the inner disk metallicity gradient inside the solar circle as earlier studies indicated. The age-metallicity relation shown by the clusters is consistent with that predicted by chemical evolution models that include the effects of radial migration, but a more detailed comparison between cluster observations and models would be premature.
△ Less
Submitted 16 May, 2016;
originally announced May 2016.
-
New Neutron-Capture Measurements in 23 Open Clusters. I. The R-Process
Authors:
Jamie C. Overbeek,
Eileen D. Friel,
Heather R. Jacobson
Abstract:
Neutron-capture elements, those with Z > 35, are the least well-understood in terms of nucleosynthesis and formation environments. The rapid neutron-capture, or r-process, elements are formed in the environments and/or remnants of massive stars, while the slow neutron-capture, or s-process, elements are primarily formed in low-mass AGB stars. These elements can provide much information about Galac…
▽ More
Neutron-capture elements, those with Z > 35, are the least well-understood in terms of nucleosynthesis and formation environments. The rapid neutron-capture, or r-process, elements are formed in the environments and/or remnants of massive stars, while the slow neutron-capture, or s-process, elements are primarily formed in low-mass AGB stars. These elements can provide much information about Galactic star formation and enrichment, but observational data is limited. We have assembled a sample of 68 stars in 23 open clusters that we use to probe abundance trends for six neutron-capture elements (Eu, Gd, Dy, Mo, Pr, and Nd) with cluster age and location in the disk of the Galaxy. In order to keep our analysis as homogenous as possible, we use an automated synthesis fitting program, which also enables us to measure multiple (3-10) lines for each element. We find that the pure r-process elements (Eu, Gd, and Dy) have positive trends with increasing cluster age, while the mixed r- and s- process elements (Mo, Pr, and Nd) have insignificant trends consistent with zero. Pr, Nd, Eu, Gd, and Dy have similar, slight (though mostly statistically significant) gradients of ~0.04 dex/kpc. The mixed elements also appear to have nonlinear relationships with Galactocentric radius.
△ Less
Submitted 19 April, 2016;
originally announced April 2016.
-
The Gaia-ESO Survey: Insights on the inner-disc evolution from open clusters
Authors:
L. Magrini,
S. Randich,
P. Donati,
A. Bragaglia,
V. Adibekyan,
D. Romano,
R. Smiljanic,
S. Blanco-Cuaresma,
G. Tautvaisiene,
E. Friel,
J. Overbeek,
H. Jacobson,
T. Cantat-Gaudin,
A. Vallenari,
R. Sordo,
E. Pancino,
D. Geisler,
I. San Roman,
S. Villanova,
A. Casey,
A. Hourihane,
C. C. Worley,
P. Francois,
G. Gilmore,
T. Bensby
, et al. (15 additional authors not shown)
Abstract:
Context. The inner disc, linking the thin disc with the bulge, has been somehow neglected in the past because of intrinsic difficulties in its study, due, e.g., to crowding and high extinction. Open clusters located in the inner disc are among the best tracers of its chemistry at different ages and distances. Aims. We analyse the chemical patterns of four open clusters located within 7 kpc of the…
▽ More
Context. The inner disc, linking the thin disc with the bulge, has been somehow neglected in the past because of intrinsic difficulties in its study, due, e.g., to crowding and high extinction. Open clusters located in the inner disc are among the best tracers of its chemistry at different ages and distances. Aims. We analyse the chemical patterns of four open clusters located within 7 kpc of the Galactic Centre and of field stars to infer the properties of the inner disc with the Gaia-ESO survey idr2/3 data release. Methods. We derive the parameters of the newly observed cluster, Berkeley 81, finding an age of about 1 Gyr and a Galactocentric distance of 5.4 kpc. We construct the chemical patterns of clusters and we compare them with those of field stars in the Solar neighbourhood and in the inner-disc samples. Results. Comparing the three populations we observe that inner-disc clusters and field stars are both, on average, enhanced in [O/Fe], [Mg/Fe] and [Si/Fe]. Using the idr2/3 results of M67, we estimate the non-local thermodynamic equilibrium (NLTE) effect on the abundances of Mg and Si in giant stars. After empirically correcting for NLTE effects, we note that NGC 6705 and Be 81 still have a high [α/Fe]. Conclusions. The location of the four open clusters and of the field population reveals that the evolution of the metallicity [Fe/H] and of [alpha/Fe] can be explained within the framework of a simple chemical evolution model: both [Fe/H] and [α/Fe] of Trumpler 20 and of NGC 4815 are in agreement with expectations from a simple chemical evolution model. On the other hand, NGC 6705, and at a lower level Berkeley 81, have higher [α/Fe] than expected for their ages, location in the disc, and metallicity. These differences might originate from local enrichment processes as explained in the inhomogeneous evolution framework.
△ Less
Submitted 15 May, 2015;
originally announced May 2015.
-
NGC 7789: An Open Cluster Case Study
Authors:
Jamie C. Overbeek,
Eileen D. Friel,
Heather R. Jacobson,
Christian I. Johnson,
Catherine A. Pilachowski,
Szabolcs Meszaros
Abstract:
We have obtained high-resolution spectra of 32 giants in the open cluster NGC 7789 using the Wisconsin-Indiana-Yale-NOAO Hydra spectrograph. We explore differences in atmospheric parameters and elemental abundances caused by the use of the linelist developed for the Gaia-ESO Survey (GES) compared to one based on Arcturus used in our previous work. [Fe/H] values decrease when using the GES linelist…
▽ More
We have obtained high-resolution spectra of 32 giants in the open cluster NGC 7789 using the Wisconsin-Indiana-Yale-NOAO Hydra spectrograph. We explore differences in atmospheric parameters and elemental abundances caused by the use of the linelist developed for the Gaia-ESO Survey (GES) compared to one based on Arcturus used in our previous work. [Fe/H] values decrease when using the GES linelist instead of the Arcturus-based linelist; these differences are probably driven by systematically lower (~ -0.1 dex) GES surface gravities. Using the GES linelist we determine abundances for 10 elements - Fe, Mg, Si, Ca, Ti, Na, Ni, Zr, Ba, and La. We find the cluster's average metallicity [Fe/H] = 0.03 +/- 0.07 dex, in good agreement with literature values, and a lower [Mg/Fe] abundance than has been reported before for this cluster (0.11 +/- 0.05 dex). We also find the neutron-capture element barium to be highly enhanced - [Ba/Fe] = +0.48 +/- 0.08 - and disparate from cluster measurements of neutron-capture elements La and Zr (-0.08 +/- 0.05 and 0.08 +/- 0.08, respectively). This is in accordance with recent discoveries of supersolar Ba enhancement in young clusters along with more modest enhancement of other neutron-capture elements formed in similar environments.
△ Less
Submitted 9 February, 2015;
originally announced February 2015.
-
Exploring Anticorrelations and Light Element Variations in Northern Globular Clusters Observed by the APOGEE Survey
Authors:
Szabolcs Meszaros,
Sarah L. Martell,
Matthew Shetrone,
Sara Lucatello,
Nicholas W. Troup,
Jo Bovy,
Katia Cunha,
Domingo A. Garcia-Hernandez,
Jamie C. Overbeek,
Carlos Allende Prieto,
Timothy C. Beers,
Peter M. Frinchaboy,
Ana E. Garcia Perez,
Fred R. Hearty,
Jon Holtzman,
Steven R. Majewski,
David L. Nidever,
Ricardo P. Schiavon,
Donald P. Schneider,
Jennifer S. Sobeck,
Verne V. Smith,
Olga Zamora,
Gail Zasowski
Abstract:
We investigate the light-element behavior of red giant stars in Northern globular clusters (GCs) observed by the SDSS-III Apache Point Observatory Galactic Evolution Experiment (APOGEE). We derive abundances of nine elements (Fe, C, N, O, Mg, Al, Si, Ca, and Ti) for 428 red giant stars in 10 globular clusters. The intrinsic abundance range relative to measurement errors is examined, and the well-k…
▽ More
We investigate the light-element behavior of red giant stars in Northern globular clusters (GCs) observed by the SDSS-III Apache Point Observatory Galactic Evolution Experiment (APOGEE). We derive abundances of nine elements (Fe, C, N, O, Mg, Al, Si, Ca, and Ti) for 428 red giant stars in 10 globular clusters. The intrinsic abundance range relative to measurement errors is examined, and the well-known C-N and Mg-Al anticorrelations are explored using an extreme-deconvolution code for the first time in a consistent way. We find that Mg and Al drive the population membership in most clusters, except in M107 and M71, the two most metal-rich clusters in our study, where the grouping is most sensitive to N. We also find a diversity in the abundance distributions, with some clusters exhibiting clear abundance bimodalities (for example M3 and M53) while others show extended distributions. The spread of Al abundances increases significantly as cluster average metallicity decreases as previously found by other works, which we take as evidence that low metallicity, intermediate mass AGB polluters were more common in the more metal poor clusters. The statistically significant correlation of [Al/Fe] with [Si/Fe] in M15 suggests that $^{28}$Si leakage has occurred in this cluster. We also present C, N and O abundances for stars cooler than 4500 K and examine the behavior of A(C+N+O) in each cluster as a function of temperature and [Al/Fe]. The scatter of A(C+N+O) is close to its estimated uncertainty in all clusters and independent on stellar temperature. A(C+N+O) exhibits small correlations and anticorrelations with [Al/Fe] in M3 and M13, but we cannot be certain about these relations given the size of our abundance uncertainties. Star-to-star variations of $α-$elements (Si, Ca, Ti) abundances are comparable to our estimated errors in all clusters.
△ Less
Submitted 4 April, 2015; v1 submitted 21 January, 2015;
originally announced January 2015.