Biomarker discovery studies for patient stratification using machine learning analysis of omics data: a scoping review

Abstract

ObjectiveTo review biomarker discovery studies using omics data for patient stratification which led to clinically validated FDA-cleared tests or laboratory developed tests, in order to identify common characteristics and derive recommendations for future biomarker projects.DesignScoping review.MethodsWe searched PubMed, EMBASE and Web of Science to obtain a comprehensive list of articles from the biomedical literature published between January 2000 and July 2021, describing clinically validated biomarker signatures for patient stratification, derived using statistical learning approaches. All documents were screened to retain only peer-reviewed research articles, review articles or opinion articles, covering supervised and unsupervised machine learning applications for omics-based patient stratification. Two reviewers independently confirmed the eligibility. Disagreements were solved by consensus. We focused the final analysis on omics-based biomarkers which achieved the highest level of validation, that is, clinical approval of the developed molecular signature as a laboratory developed test or FDA approved tests.ResultsOverall, 352 articles fulfilled the eligibility criteria. The analysis of validated biomarker signatures identified multiple common methodological and practical features that may explain the successful test development and guide future biomarker projects. These include study design choices to ensure sufficient statistical power for model building and external testing, suitable combinations of non-targeted and targeted measurement technologies, the integration of prior biological knowledge, strict filtering and inclusion/exclusion criteria, and the adequacy of statistical and machine learning methods for discovery and validation.ConclusionsWhile most clinically validated biomarker models derived from omics data have been developed for personalised oncology, first applications for non-cancer diseases show the potential of multivariate omics biomarker design for other complex disorders. Distinctive characteristics of prior success stories, such as early filtering and robust discovery approaches, continuous improvements in assay design and experimental measurement technology, and rigorous multicohort validation approaches, enable the derivation of specific recommendations for future studies.

Liquid Biopsy in Type 2 Diabetes Mellitus Management: Building Specific Biosignatures via Machine Learning

“…In addition, the accurate estimation of AUC by JADBio in small sample settings has been tested in numerous studies. Indeed, we and others have previously shown in multimodal datasets that AUC estimations did not drop upon external validation, showing no over-estimations [25][26][27][38][39][40]. This body of work proves that JADBio estimates can be trusted and there is no need to have a separate hold-out dataset to statistically validate the results, a feature of particular importance for maximal extrapolation of precious biomedical datasets.…”

“…A relatively small sample size is a limitation in this study. We [22] and others [40] argue that sample size is one of many important design elements contributing to the successful implementation in biomarker discovery. Machine learning, quickly penetrating the field, is there in order to overcome such limitations, aiding robust, optimized and maximal data extrapolation from small cohorts.…”

CKS2 and S100A12: Two Novel Diagnostic Biomarkers for Rheumatoid Arthritis

“…Importantly, the DEGs were confirmed in blood samples from RA patients. A biological marker is a patient feature evaluated as a sign for ordinary or pathologic procedure or bioresponse to therapy [ 19 , 20 ]. It has been known to us that non- or minimally invasive biological markers can be very important for the diagnosis of various diseases.…”

Study designs for clinical trials applied to personalised medicine: a scoping review

“…Methods for stratification and validation of clusters in a clinical trial (eg, data-driven subgroup identification) were considered not eligible and therefore were not included. In particular, those methods were identified and described in another recent scoping review (2021) 41. Due to the variety and diversity of trial designs currently available, this classification provides a clearer and more accessible picture of the different trial designs available in PM, helping the readers to navigate this complex field.…”