AI Science Tracker
EN

A plasma-ratio search identifies exploratory Alzheimer case-label associations

University of Massachusetts Lowell; Bedford VA Healthcare System · 28 Jan 2025 · medicine

Discovery

A plasma-ratio search identifies exploratory Alzheimer case-label associations

Source check: Source checked. Not independent scientific confirmation.

Finding: An SVM-guided search selected plasma-protein ratios associated with Alzheimer case labels in 60 stored donor samples. Candidate combinations included KYNU/MSR1, NCAN/PRTG and IL-5Rα/GFRα-1. The groups differed in average age by about 13 years, and the search reused the evaluation data; this is not independent diagnostic validation.

Technology: Linear support vector machine; feature-subset search

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Linear SVM classification and exhaustive/forward subset search select protein ratios for separating donor-group labels; the learned selection enables the represented candidate associations.
Test and result
The reanalysis used 38 control and22 Alzheimer-labeled samples, with82 protein measurements generating6,642 ordered ratios. Leave-one-out scores informed selection of best subsets; observed selected-ratio differences were then examined in the same samples.
Limitation / still unknown
Feature search and hyperparameter selection reused the LOOCV results, so headline98–100percent separation is selection-informed and not a new independent test. Mean ages were66.5 versus79.7; adding age as another classifier input does not adjust that confounding. The source also mixes22 and23 disease-sample denominators in some results. This is a reanalysis of previously collected data, without new assays, causal protein interactions or demonstrated clinical utility.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC11840181/ · https://www.ncbi.nlm.nih.gov/research/bionlp/RESTful/pmcoa.cgi/BioC_json/PMC11840181/unicode

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Official primary BioC original article sections: Clinical donor metadata;preprocessing;SVM and global feature search;ratio Results and age analysis; no laboratory-method validation.

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 1/5 · Validation 2/5 · Novelty 1/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.