Metabolic Profiles for HIV Immunological Non-Response Prediction

A cohort of 101 people living with HIV from Guangzhou Eighth People's Hospital, enrolled in late 2023, was studied to identify baseline metabolic predictors of immunological non-response after antiretroviral therapy. The dataset includes quantified plasma levels of 189 metabolites and 92 cytokines, with six metabolites showing significant differential abundance between immunological responders and non-responders. The research was authored by Heping Zhao and shared under a CC-BY-4.0 license.

Use Cases

• Train classification models to predict immunological non-response based on the six identified differential metabolites.
• Investigate correlations between specific metabolites like guanidoacetic acid and changes in CD4+ T cell counts.
• Perform feature importance analysis using random forest or SVM methods on metabolic data.
• Explore metabolic pathway dysregulation associated with poor immune recovery in HIV patients.

Strengths

• Includes data from 101 participants with clear clinical classification into 46 immunological responders and 55 non-responders.
• Quantitative analysis covers 189 plasma metabolites and 92 cytokines using UPLC-MS and Olink PEA methods.
• Provides specific model performance metrics, including a logistic regression AUC of 0.717.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small at 29.9 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Plasma metabolites and cytokines were quantified from a clinical cohort using UPLC-MS and multiplex proximity extension assays.

Time Range

Cohort established November-December 2023.

Freshness

Last updated 2026-05-15 05:44:17; freshness should be verified.

Geography

Participants were from Guangzhou Eighth People's Hospital, China.