Exploratory serum mass spectral fingerprinting and machine learning for case-control classification of preeclampsia
5 affiliations
- 1Departamento de Química, Universidad del Cauca, Campus Central, Cauca 190003 Popayán, Colombia
- 2Laboratorio de Espectroscopía Atómica y Molecular (LEAM), Universidad Industrial de Santander, Parque Tecnológico Guatiguara, Via Guatiguará Km 2, Santander 681011 Piedecuesta, Colombia
- 3Universidad Pedagógica y Tecnológica de Colombia (UPTC), Avenida Central del Norte, Boyacá 150003 Santiago de Tunja, Colombia
- 4Hospital Internacional de Colombia, Fundación Cardiovascular de Colombia, Fundación Universitaria (FCV), Valle de Menzuly Km 7, Santander 681027 Piedecuesta, Colombia
- 5Escuela de Ciencias Básicas, Facultad de Salud, Posgrado en Ciencias Biomédicas, Universidad del Valle, Campus San Fernando, Valle del Cauca 760043 Santiago de Cali, Colombia
Abstract
Background: Preeclampsia is a pregnancy-specific hypertensive disorder associated with maternal and perinatal morbidity and mortality and can increase the risk of vascular diseases after pregnancy. However, a timely and unequivocal diagnosis in the first weeks allows access to appropriate medical follow-up.
Methods: In this exploratory approach, we aim to evaluate the analytical potential of MALDI-TOF spectral fingerprints combined with machine learning algorithms for discriminating serum samples from patients with preeclampsia from those with normotensive pregnancies. The dataset comprised 164 spectra of serum samples from 67 women with preeclampsia and 97 negative controls, which were processed using the Filter-Assisted Sample Preparation (FASP) protocol and analyzed by mass spectrometry.
Results: Spectral data analysis was subjected to a machine learning algorithm that demonstrated high performance in classifying cases and controls, with an overall accuracy of 88% and sensitivities of 0.90 and 0.85 for predicting positive cases and negative controls, respectively.
Conclusions: The proposed analytical workflow demonstrated promising classification performance, highlighting the potential of MALDI-TOF spectral pattern recognition as a rapid screening approach for disease-associated signatures.
Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s12014-026-09620-5.
Related
- CT Radiomics and Machine Learning for Three-Class Bone Density Classification: An Exploratory Comparison Across Lumbar Vertebral Levels
- STCP-Net: A few-shot learning approach based on spectral-temporal convolution and an improved prototypical network for ECG classification
- Combined albumin-adjusted ischemia-modified albumin ratio and uterine artery Doppler for differentiating early-onset preeclampsia from gestational hypertension: a prospective exploratory pilot study.
- Designing an attention-based approach for RF fingerprinting in drone detection and classification
- Association of a second-trimester alpha-fetoprotein-to-transaminase ratio with preeclampsia: an exploratory retrospective cohort study.
BibTeX
@article{macacastro2026exploratory,
title = {{Exploratory serum mass spectral fingerprinting and machine learning for case-control classification of preeclampsia}},
author = {Laura Camila Maca-Castro and Luisa Fernanda Amezquita and Yuly Andrea Prada and Claudia Colmenares-Mejía and Doris Quintero-Lesmes and Paula K. Bautista-Niño and Norma C. Serrano and Enrique Mejía-Ospino},
journal = {Clinical Proteomics},
year = {2026},
month = aug,
doi = {10.1186/s12014-026-09620-5},
url = {https://doi.org/10.1186/s12014-026-09620-5},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Exploratory serum mass spectral fingerprinting and machine learning for case-control classification of preeclampsia
AU - Laura Camila Maca-Castro
AU - Luisa Fernanda Amezquita
AU - Yuly Andrea Prada
AU - Claudia Colmenares-Mejía
AU - Doris Quintero-Lesmes
AU - Paula K. Bautista-Niño
AU - Norma C. Serrano
AU - Enrique Mejía-Ospino
PY - 2026
DA - 2026/08/03
JO - Clinical Proteomics
DO - 10.1186/s12014-026-09620-5
UR - https://doi.org/10.1186/s12014-026-09620-5
AB - Background: Preeclampsia is a pregnancy-specific hypertensive disorder associated with maternal and perinatal morbidity and mortality and can increase the risk of vascular diseases after pregnancy. However, a timely and unequivocal diagnosis in the first weeks allows access to appropriate medical follow-up. Methods: In this exploratory approach, we aim to evaluate the analytical potential of MALDI-TOF spectral fingerprints combined with machine learning algorithms for discriminating serum samples from patients with preeclampsia from those with normotensive pregnancies. The dataset comprised 164 spectra of serum samples from 67 women with preeclampsia and 97 negative controls, which were processed using the Filter-Assisted Sample Preparation (FASP) protocol and analyzed by mass spectrometry. Results: Spectral data analysis was subjected to a machine learning algorithm that demonstrated high performance in classifying cases and controls, with an overall accuracy of 88% and sensitivities of 0.90 and 0.85 for predicting positive cases and negative controls, respectively. Conclusions: The proposed analytical workflow demonstrated promising classification performance, highlighting the potential of MALDI-TOF spectral pattern recognition as a rapid screening approach for disease-associated signatures. Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s12014-026-09620-5.
SN - 1542-6416
SN - 1559-0275
ER -