Human Salivary Microbiome Data for BMI Inference in Forensic Research

A pilot study by Jingyan Ji, last updated in April 2026, investigates microbial markers in human saliva for forensic body mass index (BMI) inference. The dataset likely contains results from 16S rRNA sequencing of saliva samples, with analysis identifying twenty-two microbial taxa as potential BMI indicators. A random forest model was constructed to infer BMI categories based on these microbial markers.

Use Cases

• Train machine learning models for BMI category prediction based on salivary microbial abundance data.
• Identify microbial taxa associated with different body mass index groups for biomarker discovery.
• Validate forensic investigative tools using microbiome data from biological samples like saliva.
• Conduct comparative analysis of microbial community separation between normal-weight and overweight groups.

Strengths

• Identifies twenty-two specific microbial taxa, including genera like Neisseria and Veillonella, as potential BMI-associated indicators.
• Analysis includes principal coordinates analysis showing separation between BMI groups and a constructed random forest model.
• Dataset is shared under a permissive CC-BY-4.0 license, facilitating reuse and redistribution.

Limitations

• The dataset is described as a pilot study, which may indicate a limited sample size.
• Row count and detailed column-level documentation are unknown, requiring inspection after download.
• The primary data file is in DOCX format (8.5 MB), which may not be ideal for direct computational analysis compared to structured formats like CSV.

Provenance

Source

figshare, author Jingyan Ji

Collection Method

Saliva samples were collected and subjected to 16S rRNA sequencing; microbial markers were screened using linear discriminant analysis effect size and Kruskal–Wallis rank-sum test.

Freshness

Last updated 2026-04-21 05:24:34