Tear samples are considered in recent publications as easily, noninvasively collectible information sources for precision medicine. Their complex composition may aid the identification of biomarkers and the monitoring of the effectiveness of treatments for the eye and systemic diseases. Sample collection and processing are key steps in any analytical method, especially if subtle personal differences need to be detected. In this work, we evaluate the usability of a novel sample collection technique for human tear samples using phenol red threads (cotton thread treated with the pH indicator phenol red), which are efficiently used to measure tear volume in clinical diagnosis. The low invasiveness and low discomfort to the patients have already been demonstrated, but their applicability for proteomic sample collection has not yet been compared to other methods. We have shown, using various statistical approaches, the qualitative and quantitative differences in proteomic samples collected with this novel and two traditional methods using either glass capillaries or Schirmer's paper strips. In all parameters studied, the phenol red threads proved to be equally or even more suitable than traditional methods. Based on detectability using different sampling methods, we have classified proteins in tear samples.
Keywords: LC-MS; data independent analysis; mass spectrometry; proteomics; tear.