Fig. 5: Direct profiling of clinical patient samples.

a Storage and application. The MATCH chips, each implemented with a molecular translation module and a gradient-matched signal transduction module, were lyophilized, vacuumed-packed and stored at room temperature for long-term preservation. Upon rehydration, the MATCH chips recovered with good performance. b MATCH measurements of lung cancer samples. Plasma samples collected from lung cancer patients (n = 16) and healthy controls (n = 8) were used as the lung cancer training cohort and profiled with the MATCH platform for three protein markers (EGFR, EpCAM and MUC1) and five miRNA markers (miR-21-5p, miR-30d-5p, miR-222-3p, miR-223-3p and miR-486-5p). c MATCH measurements of GBM samples. Plasma samples collected from GBM patients (n = 16) and healthy controls (n = 8) were used as GBM training cohort and profiled with the MATCH platform for three protein markers (EGFR, VEGF and EpHA2) and five miRNA markers (let-7b-5p, miR-21-5p, miR-29a-3p, miR-34a-5p, miR-223-3p). d–e Assessment of the training cohort (n = 40; 16 lung cancer samples, 16 GBM samples and 8 control samples). Through leave-one-out cross-validation, we developed regression scoring models for disease diagnosis with different combinations of biomarkers. Index scores were computed by utilizing regression models that combined the signals of protein and miRNA markers for patient classification. ***P ≤ 0.001, two-sided Student’s t test. For lung cancer, P = 0.0002. For GBM, P = 0.0002. f, g ROC analyzes of the training cohort. h, i Assessment of the validation cohort (n = 30; 10 lung cancer samples, 10 GBM samples and 10 control samples). **P ≤ 0.01, ***P ≤ 0.001, two-sided Student’s t test. For lung cancer, P = 0.0005. For GBM, P = 0.0012. j, k ROC analyzes of the validation cohort. In both lung cancer and GBM, MATCH could effectively distinguish the disease samples from the healthy controls. AUC, area under the curve. All measurements were performed in triplicate (n = 3 independent experiments) and the data are displayed as mean ± s.d. in (d, e and h, i) and as mean in (b, c). Source data are provided as a Source Data file.