Biomarker Selection Criteria Scores Model 2 for HIV and Cardiovascular Phenotype

A 5.5 KB Excel file containing results from a recursive feature addition (RFA) model for biomarker selection. The data, authored by Rachel Mac Cann and last updated in April 2026, evaluates 55 biomarkers to identify inflammatory patterns associated with cardiovascular risk in people with and without HIV. The individual additive model (Model 2) incorporated 11 additional biomarkers and produced three distinct clusters.

Use Cases

• Developing biomarker selection frameworks based on recursive feature addition methods described in the study.
• Evaluating cluster stability and biological coherence of inflammatory profiles for cardiovascular risk.
• Comparing stepwise additive versus greedy forward-backward selection models for feature inclusion.
• Analyzing associations between biomarker-derived clusters and clinical outcomes like 10-year ASCVD risk.

Strengths

• Focuses on a specific 24-marker initial panel plus 31 exploratory biomarkers for a total of 55 candidate features.
• Model 2 results are detailed, including the 11 specific additional biomarkers incorporated (e.g., GDF-15, IFN-λ2, Thrombopoietin).
• Includes clinical outcome associations, such as an adjusted odds ratio of 2.3 for cardiovascular phenotype in one cluster.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, author Rachel Mac Cann.

Collection Method

Derived from a research study applying recursive feature addition models to biomarker data.

Time Range

null

Freshness

Last updated 2026-04-27 17:33:29; freshness should be verified.

Geography

null