Computational biomarker pipeline from discovery to clinical implementation: plasma proteomic biomarkers for cardiac transplantation

Abstract

Recent technical advances in the field of quantitative proteomics have stimulated a large number of biomarker discovery studies of various diseases, providing avenues for new treatments and diagnostics. However, inherent challenges have limited the successful translation of candidate biomarkers into clinical use, thus highlighting the need for a robust analytical methodology to transition from biomarker discovery to clinical implementation. We have developed an end-to-end computational proteomic pipeline for biomarkers studies. At the discovery stage, the pipeline emphasizes different aspects of experimental design, appropriate statistical methodologies, and quality assessment of results. At the validation stage, the pipeline focuses on the migration of the results to a platform appropriate for external validation, and the development of a classifier score based on corroborated protein biomarkers. At the last stage towards clinical implementation, the main aims are to develop and validate an assay suitable for clinical deployment, and to calibrate the biomarker classifier using the developed assay. The proposed pipeline was applied to a biomarker study in cardiac transplantation aimed at developing a minimally invasive clinical test to monitor acute rejection. Starting with an untargeted screening of the human plasma proteome, five candidate biomarker proteins were identified. Rejection-regulated proteins reflect cellular and humoral immune responses, acute phase inflammatory pathways, and lipid metabolism biological processes. A multiplex multiple reaction monitoring mass-spectrometry (MRM-MS) assay was developed for the five candidate biomarkers and validated by enzyme-linked immune-sorbent (ELISA) and immunonephelometric assays (INA). A classifier score based on corroborated proteins demonstrated that the developed MRM-MS assay provides an appropriate methodology for an external validation, which is still in progress. Plasma proteomic biomarkers of acute cardiac rejection may offer a relevant post-transplant monitoring tool to effectively guide clinical care. The proposed computational pipeline is highly applicable to a wide range of biomarker proteomic studies.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Figure 1. Proteomic computational pipeline synopsis.

The 3-stage computational pipeline enables an initial untargeted exploration of the plasma proteome resulting in a list of potential biomarkers, followed by the validation of a set of candidate biomarkers that emphasizes the combination of candidate protein biomarkers into a classifier score with clinical utility. The bottom panel outlines the main steps of the computational pipeline that provide a systematic process from discovery to validation to clinical implementation of plasma protein biomarkers.

Figure 2. Plasma protein panel.

Average Linear Discriminant Analysis classifier score (Classifier score) for all available acute rejection (AR) samples (pink solid point), and non-rejection (NR) samples from NR patients (green solid triangle), at each time point. The score was centered at the LDA cut-off point so that samples with positive and negative scores are classified as “rejection” or “non-rejection”, respectively. Vertical lines represent standard errors. Means and standard error bars can be used to assess differences of the score between groups at any of the studied time points. Sample sizes available at each time point are shown in the bottom table.

Figure 3. Transition plot.

Linear Discriminant Analysis classifier score (Classifier score) when patients transitioned between non-rejection (NR) and acute rejection (AR) episodes. The first consecutive AR time points were averaged (AR, pink solid point) from 7 AR patients. Non-rejection samples from the same patients, before and after AR (“NR before AR” and “NR after AR”) were averaged (pink solid triangle). The average time trend for these samples is represented with a pink solid line. A control curve (dashed green line) was constructed from 9 NR patients matched to AR patients by available time points (green solid triangle). Vertical lines represent standard errors. The asterisk (*) means that the two-sided t-test p value<0.001.

Figure 4. Technical validation.

A. For both ELISA/INA and MRM-MS corroboration analysis, p values were calculated by robust eBayes (12 acute rejection (AR) versus 18 non-rejection (NR) samples in ELISA/INA, and 6 AR versus 11 NR in MRM-MS, two-sided test). The correlations among platforms are based on all available common samples, i.e., 25 samples measured by iTRAQ and ELISA/INA, and 23 samples measured by iTRAQ, ELISA/INA, and MRM-MS (Figure S2). B. Validation performance (y-axis) estimated by a 6-fold cross-validation: sensitivity (blue diamond), specificity (brown square), and area under the receiver operating curve (AUC) (red star) for incremental classifier panels. The sensitivity and specificity estimates were calculated using a probability cut-off of 0.5. The x-axis shows three nested classifier panels based on a single candidate marker (B2M), 2 markers (B2M&ADIPOQ) and 3 markers (B2M&ADIPOQ&CP), respectively, measured by MRM-MS. As F10 was not validated in either ELISA/INA or MRM-MS, it was not included in any MRM-based classifier. C. MRM-MS classifier score generated by a 6-fold cross-validation using Linear Discriminant Analysis. Samples with a positive proteomic classifier score are classified as “rejection” and those with a negative score are classified as “non-rejection”.