Extracting meaningful features from time series analysis of plasma analytes in clinical research
- 1NICHD, NIH
Abstract
While absolute blood biomarkers have long guided diagnosis, their rate of change often provides greater diagnostic value. For the Limb Girdle Muscular Dystrophies studied here, increased muscle weakness is associated with abnormal sarcolemma repair. Time course analyses were made to changes in muscle-derived plasma enzymes after modest physical activity of ambulatory inpatients. Blood samples at 20 selected time points at variable intervals were obtained at rest, and before and after routine morning activities and motor function testing. After these activities, plasma enzyme levels increased in a similar way, suggesting damage and leakage from skeletal muscle. The levels then decreased in a similar way, suggesting that muscle cells stopped leaking and enzymes were subsequently cleared from the blood. The complex temporal changes were modeled using both continuous time and data features to reduce the number of parameters required to describe the irregularly sampled, interrupted, time series data. Unlike individuals with cardiac muscle ischemia or following intense exercise, there was no extended post-injury rising phase of enzyme levels. In contradistinction, levels fell once activity ceased. Time series analyses of blood biomarkers most likely provide an assay and index of sarcolemma fragility and by extension skeletal muscle damage.
Related
- Clinical and genetic features, surgical management, and longitudinal follow-up in two children with progressive osseous heteroplasia
- Clinical and MEFV variant characteristics of pediatric patients with an FMF-compatible phenotype and PFAPA-compatible features in a Palestinian referral cohort: a retrospective study
- Clinical Features, Genetics and Treatment in a Russian Cohort of Young Patients with Xeroderma Pigmentosum: A Single Center Experience
- Multi-cancer screening in peritoneal dialysis patients: an explainable machine learning framework integrating bioelectrical impedance and clinical features
- Hereditary transthyretin amyloidosis in peripheral neuropathy: prevalence, clinical features, and neurology-based screening in non-endemic Japan
BibTeX
@misc{blank2026extracting,
title = {{Extracting meaningful features from time series analysis of plasma analytes in clinical research}},
author = {P. Blank and A. E. Golding and I. Morales and H. Waters and E. Mekhedov and L. Bezrukov and R. D. Wachter and I. Mikhailenko and R. H. Brown and C. Bonnemann and A. Demidowich and M. Jain and J. A. Yanovski and J. Zimmerberg},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.08.26365091},
url = {https://www.medrxiv.org/content/10.64898/2026.10.08.26365091v1},
note = {CC0},
}
RIS
TY - UNPB
TI - Extracting meaningful features from time series analysis of plasma analytes in clinical research
AU - P. Blank
AU - A. E. Golding
AU - I. Morales
AU - H. Waters
AU - E. Mekhedov
AU - L. Bezrukov
AU - R. D. Wachter
AU - I. Mikhailenko
AU - R. H. Brown
AU - C. Bonnemann
AU - A. Demidowich
AU - M. Jain
AU - J. A. Yanovski
AU - J. Zimmerberg
PY - 2026
DA - 2026/10/08
PB - medRxiv
DO - 10.64898/2026.10.08.26365091
UR - https://www.medrxiv.org/content/10.64898/2026.10.08.26365091v1
AB - While absolute blood biomarkers have long guided diagnosis, their rate of change often provides greater diagnostic value. For the Limb Girdle Muscular Dystrophies studied here, increased muscle weakness is associated with abnormal sarcolemma repair. Time course analyses were made to changes in muscle-derived plasma enzymes after modest physical activity of ambulatory inpatients. Blood samples at 20 selected time points at variable intervals were obtained at rest, and before and after routine morning activities and motor function testing. After these activities, plasma enzyme levels increased in a similar way, suggesting damage and leakage from skeletal muscle. The levels then decreased in a similar way, suggesting that muscle cells stopped leaking and enzymes were subsequently cleared from the blood. The complex temporal changes were modeled using both continuous time and data features to reduce the number of parameters required to describe the irregularly sampled, interrupted, time series data. Unlike individuals with cardiac muscle ischemia or following intense exercise, there was no extended post-injury rising phase of enzyme levels. In contradistinction, levels fell once activity ceased. Time series analyses of blood biomarkers most likely provide an assay and index of sarcolemma fragility and by extension skeletal muscle damage.
ER -