An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients

Gustavo Martinez; Alexis Garduno; Abdullah Mahmud-Al-Rafat; Ali Toloue Ostadgavahi; Ann Avery; Scheila de Avila e Silva; Rachael Cusack; Cheryl Cameron; Mark Cameron; Ignacio Martin-Loeches; David Kelvin

doi:10.7717/peerj.14487

An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients

Gustavo Martinez, Alexis Garduno, Abdullah Mahmud-Al-Rafat, Ali Toloue Ostadgavahi, Ann Avery, Scheila de Avila e Silva, Rachael Cusack, Cheryl Cameron, Mark Cameron, Ignacio Martin-Loeches, David Kelvin

Medicine

Producción científica: Contribución a una revista › Artículo › revisión exhaustiva

3 Citas (Scopus)

Resumen

Background. The severe form of COVID-19 can cause a dysregulated host immune syndrome that might lead patients to death. To understand the underlying immune mechanisms that contribute to COVID-19 disease we have examined 28 different biomarkers in two cohorts of COVID-19 patients, aiming to systematically capture, quantify, and algorithmize how immune signals might be associated to the clinical outcome of COVID-19 patients. Methods. The longitudinal concentration of 28 biomarkers of 95 COVID-19 patients was measured. We performed a dimensionality reduction analysis to determine meaningful biomarkers for explaining the data variability. The biomarkers were used as input of artificial neural network, random forest, classification and regression trees, k-nearest neighbors and support vector machines. Two different clinical cohorts were used to grant validity to the findings. Results. We benchmarked the classification capacity of two COVID-19 clinicals studies with different models and found that artificial neural networks was the best classifier. From it, we could employ different sets of biomarkers to predict the clinical outcome of COVID-19 patients. First, all the biomarkers available yielded a satisfactory classification. Next, we assessed the prediction capacity of each protein separated. With a reduced set of biomarkers, our model presented 94% accuracy, 96.6% precision, 91.6% recall, and 95% of specificity upon the testing data. We used the same model to predict 83% and 87% (recovered and deceased) of unseen data, granting validity to the results obtained. Conclusions. In this work, using state-of-the-art computational techniques, we systematically identified an optimal set of biomarkers that are related to a prediction capacity of COVID-19 patients. The screening of such biomarkers might assist in understanding the underlying immune response towards inflammatory diseases.

Idioma original	English
Número de artículo	e14487
Publicación	PeerJ
Volumen	10
DOI	https://doi.org/10.7717/peerj.14487
Estado	Published - dic. 2022

Nota bibliográfica

Funding Information:
We thank the Canadian Research Chair in Translational Vaccinology and Inflammation. We thank Michelle Hecker and Claudia Ute Sontich from MetroHealth Medical Center and Banumathi Tamilselvan from Case Western Reserve University for their contributions in patient coordination and sample acquisition. We thank all the members from the Laboratory of Emerging and Infectious Diseases at Dalhousie University. Finally, we also thank Nikki Kelvin for her key inputs to this manuscript. This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Funding Information:
This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Publisher Copyright:
Copyright 2022 Martinez et al.

ASJC Scopus Subject Areas

General Neuroscience
General Biochemistry,Genetics and Molecular Biology
General Agricultural and Biological Sciences

PubMed: MeSH publication types

Journal Article
Research Support, Non-U.S. Gov't

Acceder al documento

10.7717/peerj.14487

Otros archivos y enlaces

Citar esto

Martinez, G., Garduno, A., Mahmud-Al-Rafat, A., Ostadgavahi, A. T., Avery, A., de Avila e Silva, S., Cusack, R., Cameron, C., Cameron, M., Martin-Loeches, I., & Kelvin, D. (2022). An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients. PeerJ, 10, Artículo e14487. https://doi.org/10.7717/peerj.14487

An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients. / Martinez, Gustavo; Garduno, Alexis; Mahmud-Al-Rafat, Abdullah et al.
En: PeerJ, Vol. 10, e14487, 12.2022.

Producción científica: Contribución a una revista › Artículo › revisión exhaustiva

Martinez, G, Garduno, A, Mahmud-Al-Rafat, A, Ostadgavahi, AT, Avery, A, de Avila e Silva, S, Cusack, R, Cameron, C, Cameron, M, Martin-Loeches, I & Kelvin, D 2022, 'An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients', PeerJ, vol. 10, e14487. https://doi.org/10.7717/peerj.14487

@article{854b38a639154fd79f922816d3f55e7e,

title = "An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients",

abstract = "Background. The severe form of COVID-19 can cause a dysregulated host immune syndrome that might lead patients to death. To understand the underlying immune mechanisms that contribute to COVID-19 disease we have examined 28 different biomarkers in two cohorts of COVID-19 patients, aiming to systematically capture, quantify, and algorithmize how immune signals might be associated to the clinical outcome of COVID-19 patients. Methods. The longitudinal concentration of 28 biomarkers of 95 COVID-19 patients was measured. We performed a dimensionality reduction analysis to determine meaningful biomarkers for explaining the data variability. The biomarkers were used as input of artificial neural network, random forest, classification and regression trees, k-nearest neighbors and support vector machines. Two different clinical cohorts were used to grant validity to the findings. Results. We benchmarked the classification capacity of two COVID-19 clinicals studies with different models and found that artificial neural networks was the best classifier. From it, we could employ different sets of biomarkers to predict the clinical outcome of COVID-19 patients. First, all the biomarkers available yielded a satisfactory classification. Next, we assessed the prediction capacity of each protein separated. With a reduced set of biomarkers, our model presented 94% accuracy, 96.6% precision, 91.6% recall, and 95% of specificity upon the testing data. We used the same model to predict 83% and 87% (recovered and deceased) of unseen data, granting validity to the results obtained. Conclusions. In this work, using state-of-the-art computational techniques, we systematically identified an optimal set of biomarkers that are related to a prediction capacity of COVID-19 patients. The screening of such biomarkers might assist in understanding the underlying immune response towards inflammatory diseases.",

author = "Gustavo Martinez and Alexis Garduno and Abdullah Mahmud-Al-Rafat and Ostadgavahi, {Ali Toloue} and Ann Avery and {de Avila e Silva}, Scheila and Rachael Cusack and Cheryl Cameron and Mark Cameron and Ignacio Martin-Loeches and David Kelvin",

note = "Funding Information: We thank the Canadian Research Chair in Translational Vaccinology and Inflammation. We thank Michelle Hecker and Claudia Ute Sontich from MetroHealth Medical Center and Banumathi Tamilselvan from Case Western Reserve University for their contributions in patient coordination and sample acquisition. We thank all the members from the Laboratory of Emerging and Infectious Diseases at Dalhousie University. Finally, we also thank Nikki Kelvin for her key inputs to this manuscript. This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Funding Information: This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Publisher Copyright: Copyright 2022 Martinez et al.",

year = "2022",

month = dec,

doi = "10.7717/peerj.14487",

language = "English",

volume = "10",

journal = "PeerJ",

issn = "2167-8359",

publisher = "PeerJ",

}

TY - JOUR

T1 - An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients

AU - Martinez, Gustavo

AU - Garduno, Alexis

AU - Mahmud-Al-Rafat, Abdullah

AU - Ostadgavahi, Ali Toloue

AU - Avery, Ann

AU - de Avila e Silva, Scheila

AU - Cusack, Rachael

AU - Cameron, Cheryl

AU - Cameron, Mark

AU - Martin-Loeches, Ignacio

AU - Kelvin, David

N1 - Funding Information: We thank the Canadian Research Chair in Translational Vaccinology and Inflammation. We thank Michelle Hecker and Claudia Ute Sontich from MetroHealth Medical Center and Banumathi Tamilselvan from Case Western Reserve University for their contributions in patient coordination and sample acquisition. We thank all the members from the Laboratory of Emerging and Infectious Diseases at Dalhousie University. Finally, we also thank Nikki Kelvin for her key inputs to this manuscript. This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Funding Information: This work was supported by awards from the Canadian Institutes of Health Research, the Canadian 2019 Novel Coronavirus (COVID-19) Rapid Research Funding initiative (CIHR OV2 –170357), Research Nova Scotia (David Kelvin), Atlantic Genome/Genome Canada (David Kelvin), Li-Ka Shing Foundation (David Kelvin), and the Dalhousie Medical Research Foundation (David Kelvin). This study has also been supported by SFI (Science Foundation Ireland), Grant Number 20/COV/0038 (IML). This work was also supported by a donation from the Nord Family Foundation to Mark Cameron, Ann Avery, and Cheryl Cameron. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Publisher Copyright: Copyright 2022 Martinez et al.

PY - 2022/12

Y1 - 2022/12

N2 - Background. The severe form of COVID-19 can cause a dysregulated host immune syndrome that might lead patients to death. To understand the underlying immune mechanisms that contribute to COVID-19 disease we have examined 28 different biomarkers in two cohorts of COVID-19 patients, aiming to systematically capture, quantify, and algorithmize how immune signals might be associated to the clinical outcome of COVID-19 patients. Methods. The longitudinal concentration of 28 biomarkers of 95 COVID-19 patients was measured. We performed a dimensionality reduction analysis to determine meaningful biomarkers for explaining the data variability. The biomarkers were used as input of artificial neural network, random forest, classification and regression trees, k-nearest neighbors and support vector machines. Two different clinical cohorts were used to grant validity to the findings. Results. We benchmarked the classification capacity of two COVID-19 clinicals studies with different models and found that artificial neural networks was the best classifier. From it, we could employ different sets of biomarkers to predict the clinical outcome of COVID-19 patients. First, all the biomarkers available yielded a satisfactory classification. Next, we assessed the prediction capacity of each protein separated. With a reduced set of biomarkers, our model presented 94% accuracy, 96.6% precision, 91.6% recall, and 95% of specificity upon the testing data. We used the same model to predict 83% and 87% (recovered and deceased) of unseen data, granting validity to the results obtained. Conclusions. In this work, using state-of-the-art computational techniques, we systematically identified an optimal set of biomarkers that are related to a prediction capacity of COVID-19 patients. The screening of such biomarkers might assist in understanding the underlying immune response towards inflammatory diseases.

AB - Background. The severe form of COVID-19 can cause a dysregulated host immune syndrome that might lead patients to death. To understand the underlying immune mechanisms that contribute to COVID-19 disease we have examined 28 different biomarkers in two cohorts of COVID-19 patients, aiming to systematically capture, quantify, and algorithmize how immune signals might be associated to the clinical outcome of COVID-19 patients. Methods. The longitudinal concentration of 28 biomarkers of 95 COVID-19 patients was measured. We performed a dimensionality reduction analysis to determine meaningful biomarkers for explaining the data variability. The biomarkers were used as input of artificial neural network, random forest, classification and regression trees, k-nearest neighbors and support vector machines. Two different clinical cohorts were used to grant validity to the findings. Results. We benchmarked the classification capacity of two COVID-19 clinicals studies with different models and found that artificial neural networks was the best classifier. From it, we could employ different sets of biomarkers to predict the clinical outcome of COVID-19 patients. First, all the biomarkers available yielded a satisfactory classification. Next, we assessed the prediction capacity of each protein separated. With a reduced set of biomarkers, our model presented 94% accuracy, 96.6% precision, 91.6% recall, and 95% of specificity upon the testing data. We used the same model to predict 83% and 87% (recovered and deceased) of unseen data, granting validity to the results obtained. Conclusions. In this work, using state-of-the-art computational techniques, we systematically identified an optimal set of biomarkers that are related to a prediction capacity of COVID-19 patients. The screening of such biomarkers might assist in understanding the underlying immune response towards inflammatory diseases.

UR - http://www.scopus.com/inward/record.url?scp=85145369816&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85145369816&partnerID=8YFLogxK

U2 - 10.7717/peerj.14487

DO - 10.7717/peerj.14487

M3 - Article

C2 - 36530391

AN - SCOPUS:85145369816

SN - 2167-8359

VL - 10

JO - PeerJ

JF - PeerJ

M1 - e14487

ER -

An artificial neural network classification method employing longitudinally monitored immune biomarkers to predict the clinical outcome of critically ill COVID-19 patients

Resumen

Nota bibliográfica

ASJC Scopus Subject Areas

PubMed: MeSH publication types

Acceder al documento

Otros archivos y enlaces

Huella

Citar esto