About this resource
<p><span>Adoption of high-content omic technologies in clinical studies, coupled with computational </span><span>methods, have yielded an abundance of candidate biomarkers. However, translating such find</span><span>ings into bona fide clinical biomarkers remains challenging.</span> <span>To facilitate this process, we </span><span>introduce Stabl, a general machine learning framework that identifies a sparse, reliable set </span><span>of biomarkers by integrating noise injection and a data-driven signal-to-noise threshold into </span><span>multivariable predictive modeling.</span> <span>Evaluation of Stabl on synthetic datasets and five inde</span><span>pendent clinical studies demonstrates improved biomarker sparsity and reliability compared to </span><span>commonly used sparsity-promoting regularization methods while maintaining predictive per</span><span>formance; it distills datasets containing 1,400 to 35,000 features down to 4 to 34 candidate </span><span>biomarkers. Stabl extends to multi-omic integration tasks, enabling biological interpretation of </span><span>complex predictive models, as it hones in on a shortlist of proteomic, metabolomic, and cyto</span><span>metric events predicting labor onset, microbial biomarkers of preterm birth, and a pre-operative </span><span>immune signature of post-surgical infections.</span></p>, <p>Funding provided by: Stanford University<br>Crossref Funder Registry ID: http://dx.doi.org/10.13039/100005492<br>Award Number: </p>
Files and access
Resources are read online through DARE. Direct file links are not published; access is checked against DSpace when a document is opened. Authorised librarian workflows are unaffected.
This record contains bibliographic metadata, but no files are currently attached.
