SERS Spectral Dataset of Filtered and Unfiltered Plasma Samples for Latent Tuberculosis Infection Prediction

This dataset accompanies the manuscript titled “Filtering Large Plasma Proteins Enhances Raman Spectroscopic Signal Detection and Diagnostic Accuracy for Latent Tuberculosis Infection.” It contains SERS spectra from plasma samples classified as interferon-gamma release assay (IGRA) positive and negative. The plasma samples were prepared under both filtered and unfiltered conditions. The dataset was used for spectral preprocessing, machine learning–based analysis, and model validation in tuberculosis prediction studies.