Identifying Depression Through Machine Learning Analysis of Omics Data: Scoping Review.
Brittany Taylor, Mollie Hobensack, Stephanie Niño de Rivera and 3 others
PMID 39028994WHAT IT FOUND
A scoping review of 15 studies found that machine learning models using blood or saliva DNA can predict depression, but results vary widely and lack reproducibility.
This technology is not ready for clinical diagnosis. Nurses cannot use these omics scores to screen patients today.
Key findings
01The review included 15 cross-sectional studies using machine learning to analyze omics data (genomics, transcriptomics, epigenomics, microbiomics, multiomics) to identify depression.
02Study results showed minimal overlap and low reproducibility, indicating the field is still in exploratory stages rather than ready for clinical implementation.
03None of the 15 included studies had nurse researchers on their study teams.
STILL TO COME
How it was doneWhat they foundWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
All included studies were cross-sectional, so they cannot prove that omics data cause or predict depression over time. There was significant heterogeneity in methods and metrics, making it impossible to compare performance across studies. Many studies used biobank data with poor participant descriptions, limiting generalizability. 47% of studies did not report how depression was diagnosed or screened. The review did not include cost-effectiveness analyses, which are crucial for clinical implementation.
Declared interests
None declared.
The easy way to misread this
Do not interpret the high accuracy scores reported in individual studies (e.g., AUC >0.90) as evidence that machine learning depression diagnosis is ready for clinical practice. The review explicitly states that results lack reproducibility and overlap, and the field is still in exploratory stages.